Use CentOS 9 instead of 8

* Make `IfOperator` return the `else_expression_` in case of `NULL`

* Add gql_behave tests

* Add gql_behave test to specifically check for the case when the test expression itself is null

.clang-tidy

@@ -61,7 +61,9 @@ Checks: '*,
          -readability-magic-numbers,
          -readability-named-parameter,
          -misc-no-recursion,
-         -concurrency-mt-unsafe'
+         -concurrency-mt-unsafe,
+         -bugprone-easily-swappable-parameters'
+
 WarningsAsErrors: ''
 HeaderFilterRegex: 'src/.*'
 AnalyzeTemporaryDtors: false
@@ -88,4 +90,3 @@ CheckOptions:
   - key:             modernize-use-nullptr.NullMacros
     value:           'NULL'
 ...
-

.githooks/pre-commit

@@ -24,14 +24,6 @@ for file in $modified_files; do
 
   git checkout-index --prefix="$tmpdir/" -- $file
 
-  echo "Running clang-format..."
-  $project_folder/tools/git-clang-format $tmpdir/$file
-  CODE=$?
-
-  if [ $CODE -ne 0 ]; then
-    break
-  fi
-
   # Do not break header checker
   echo "Running header checker..."
   $project_folder/tools/header-checker.py $tmpdir/$file $file --amend-year
@@ -39,7 +31,6 @@ for file in $modified_files; do
   if [ $CODE -ne 0 ]; then
     FAIL=1
   fi
-
 done;
 
 return ${FAIL}

.github/workflows/diff.yaml

@@ -1,4 +1,7 @@
 name: Diff
+concurrency:
+  group: ${{ github.head_ref || github.sha }}
+  cancel-in-progress: true
 
 on:
   push:
@@ -112,7 +115,7 @@ jobs:
           source /opt/toolchain-v4/activate
 
           # Restrict clang-tidy results only to the modified parts
-          git diff -U0 master... -- src ':!*.hpp' | ./tools/github/clang-tidy/clang-tidy-diff.py -p 1 -j $THREADS -path build | tee ./build/clang_tidy_output.txt
+          git diff -U0 master... -- src | ./tools/github/clang-tidy/clang-tidy-diff.py -p 1 -j $THREADS -path build | tee ./build/clang_tidy_output.txt
 
           # Fail if any warning is reported
           ! cat ./build/clang_tidy_output.txt | ./tools/github/clang-tidy/grep_error_lines.sh > /dev/null

.github/workflows/package_all.yaml

@@ -6,11 +6,11 @@ on: workflow_dispatch
 
 jobs:
   centos-7:
-    runs-on: [self-hosted, DockerMgBuild]
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -22,29 +22,29 @@ jobs:
           name: centos-7
           path: build/output/centos-7/memgraph*.rpm
 
-  centos-8:
-    runs-on: [self-hosted, DockerMgBuild]
+  centos-9:
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
         run: |
-          ./release/package/run.sh package centos-8
+          ./release/package/run.sh package centos-9
       - name: "Upload package"
         uses: actions/upload-artifact@v2
         with:
-          name: centos-8
-          path: build/output/centos-8/memgraph*.rpm
+          name: centos-9
+          path: build/output/centos-9/memgraph*.rpm
 
   debian-10:
-    runs-on: [self-hosted, DockerMgBuild]
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -57,11 +57,11 @@ jobs:
           path: build/output/debian-10/memgraph*.deb
 
   debian-11:
-    runs-on: [self-hosted, DockerMgBuild]
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -74,11 +74,11 @@ jobs:
           path: build/output/debian-11/memgraph*.deb
 
   docker:
-    runs-on: [self-hosted, DockerMgBuild]
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -93,11 +93,11 @@ jobs:
           path: build/output/docker/memgraph*.tar.gz
 
   ubuntu-1804:
-    runs-on: [self-hosted, DockerMgBuild]
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -110,11 +110,11 @@ jobs:
           path: build/output/ubuntu-18.04/memgraph*.deb
 
   ubuntu-2004:
-    runs-on: [self-hosted, DockerMgBuild]
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -126,12 +126,29 @@ jobs:
           name: ubuntu-2004
           path: build/output/ubuntu-20.04/memgraph*.deb
 
-  debian-11-platform:
-    runs-on: [self-hosted, DockerMgBuild]
+  ubuntu-2204:
+    runs-on: [self-hosted, DockerMgBuild, X64]
     timeout-minutes: 60
     steps:
       - name: "Set up repository"
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
+        with:
+          fetch-depth: 0 # Required because of release/get_version.py
+      - name: "Build package"
+        run: |
+          ./release/package/run.sh package ubuntu-22.04
+      - name: "Upload package"
+        uses: actions/upload-artifact@v2
+        with:
+          name: ubuntu-2204
+          path: build/output/ubuntu-22.04/memgraph*.deb
+
+  debian-11-platform:
+    runs-on: [self-hosted, DockerMgBuild, X64]
+    timeout-minutes: 60
+    steps:
+      - name: "Set up repository"
+        uses: actions/checkout@v3
         with:
           fetch-depth: 0 # Required because of release/get_version.py
       - name: "Build package"
@@ -142,3 +159,20 @@ jobs:
         with:
           name: debian-11-platform
           path: build/output/debian-11/memgraph*.deb
+
+  debian-11-arm:
+    runs-on: [self-hosted, DockerMgBuild, ARM64, strange]
+    timeout-minutes: 60
+    steps:
+      - name: "Set up repository"
+        uses: actions/checkout@v3
+        with:
+          fetch-depth: 0 # Required because of release/get_version.py
+      - name: "Build package"
+        run: |
+          ./release/package/run.sh package debian-11-arm
+      - name: "Upload package"
+        uses: actions/upload-artifact@v2
+        with:
+          name: debian-11-arm
+          path: build/output/debian-11-arm/memgraph*.deb

.github/workflows/release_centos9.yaml

@@ -1,4 +1,4 @@
-name: Release CentOS 8
+name: Release CentOS 9
 
 on:
   workflow_dispatch:
@@ -8,7 +8,7 @@ on:
 jobs:
   community_build:
     name: "Community build"
-    runs-on: [self-hosted, Linux, X64, CentOS8]
+    runs-on: [self-hosted, Linux, X64, CentOS9]
     env:
       THREADS: 24
       MEMGRAPH_ENTERPRISE_LICENSE: ${{ secrets.MEMGRAPH_ENTERPRISE_LICENSE }}
@@ -47,7 +47,7 @@ jobs:
 
   coverage_build:
     name: "Coverage build"
-    runs-on: [self-hosted, Linux, X64, CentOS8]
+    runs-on: [self-hosted, Linux, X64, CentOS9]
     env:
       THREADS: 24
       MEMGRAPH_ENTERPRISE_LICENSE: ${{ secrets.MEMGRAPH_ENTERPRISE_LICENSE }}
@@ -104,7 +104,7 @@ jobs:
 
   debug_build:
     name: "Debug build"
-    runs-on: [self-hosted, Linux, X64, CentOS8]
+    runs-on: [self-hosted, Linux, X64, CentOS9]
     env:
       THREADS: 24
       MEMGRAPH_ENTERPRISE_LICENSE: ${{ secrets.MEMGRAPH_ENTERPRISE_LICENSE }}
@@ -180,7 +180,7 @@ jobs:
 
   release_build:
     name: "Release build"
-    runs-on: [self-hosted, Linux, X64, CentOS8]
+    runs-on: [self-hosted, Linux, X64, CentOS9]
     env:
       THREADS: 24
       MEMGRAPH_ENTERPRISE_LICENSE: ${{ secrets.MEMGRAPH_ENTERPRISE_LICENSE }}

.github/workflows/release_docker.yaml

@@ -4,8 +4,12 @@ on:
   workflow_dispatch:
     inputs:
       version:
-        description: "Memgraph binary version to publish on Dockerhub."
+        description: "Memgraph binary version to publish on DockerHub."
         required: true
+      force_release:
+        type: boolean
+        required: false
+        default: false
 
 jobs:
   docker_publish:
@@ -36,6 +40,22 @@ jobs:
           curl -L https://download.memgraph.com/memgraph/v${{ github.event.inputs.version }}/debian-11/memgraph_${{ github.event.inputs.version }}-1_amd64.deb > memgraph-amd64.deb
           curl -L https://download.memgraph.com/memgraph/v${{ github.event.inputs.version }}/debian-11-aarch64/memgraph_${{ github.event.inputs.version }}-1_arm64.deb > memgraph-arm64.deb
 
+      - name: Check if specified version is already pushed
+        run: |
+          EXISTS=$(docker manifest inspect $DOCKER_ORGANIZATION_NAME/$DOCKER_REPOSITORY_NAME:${{ github.event.inputs.version }} > /dev/null; echo $?)
+          echo $EXISTS
+          if [[ ${EXISTS} -eq 0 ]]; then
+              echo 'The specified version has been already released to DockerHub.'
+              if [[ ${{ github.event.inputs.force_release }} = true ]]; then
+                echo 'Forcing the release!'
+              else
+                echo 'Stopping the release!'
+                exit 1
+              fi
+          else
+              echo 'All good the specified version has not been release to DockerHub.'
+          fi
+
       - name: Build & push docker images
         run: |
           cd release/docker

.pre-commit-config.yaml

@@ -0,0 +1,24 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v2.3.0
+    hooks:
+      - id: check-yaml
+      - id: end-of-file-fixer
+      - id: trailing-whitespace
+  - repo: https://github.com/psf/black
+    rev: 22.3.0
+    hooks:
+      - id: black
+        args: # arguments to configure black
+          - --line-length=120
+          - --include='\.pyi?$'
+          # these folders wont be formatted by black
+          - --exclude="""\.git |
+            \.__pycache__|
+            build|
+            libs|
+            .cache"""
+  - repo: https://github.com/pre-commit/mirrors-clang-format
+    rev: v13.0.0
+    hooks:
+      - id: clang-format

CODEOWNERS

@@ -1 +0,0 @@
-* @antaljanosbenjamin @kostasrim

config/flags.yaml

@@ -47,6 +47,14 @@ modifications:
     value: ""
     override: false
 
+  - name: "bolt_cert_file"
+    value: "/etc/memgraph/ssl/cert.pem"
+    override: false
+
+  - name: "bolt_key_file"
+    value: "/etc/memgraph/ssl/key.pem"
+    override: false
+
   - name: "storage_properties_on_edges"
     value: "true"
     override: true

environment/os/centos-8.sh

@@ -1,161 +0,0 @@
-#!/bin/bash
-
-set -Eeuo pipefail
-
-DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
-source "$DIR/../util.sh"
-
-TOOLCHAIN_BUILD_DEPS=(
-    coreutils-common gcc gcc-c++ make # generic build tools
-    wget # used for archive download
-    gnupg2 # used for archive signature verification
-    tar gzip bzip2 xz unzip # used for archive unpacking
-    zlib-devel # zlib library used for all builds
-    expat-devel libipt libipt-devel libbabeltrace-devel xz-devel python36-devel texinfo # for gdb
-    libcurl-devel # for cmake
-    curl # snappy
-    readline-devel # for cmake and llvm
-    libffi-devel libxml2-devel # for llvm
-    libedit-devel pcre-devel automake bison # for swig
-    file
-    openssl-devel
-    gmp-devel
-    gperf
-    patch
-)
-
-TOOLCHAIN_RUN_DEPS=(
-    make # generic build tools
-    tar gzip bzip2 xz # used for archive unpacking
-    zlib # zlib library used for all builds
-    expat libipt libbabeltrace xz-libs python36 # for gdb
-    readline # for cmake and llvm
-    libffi libxml2 # for llvm
-    openssl-devel
-)
-
-MEMGRAPH_BUILD_DEPS=(
-    git # source code control
-    make pkgconf-pkg-config # build system
-    curl wget # for downloading libs
-    libuuid-devel java-11-openjdk # required by antlr
-    readline-devel # for memgraph console
-    python36-devel # for query modules
-    openssl-devel
-    libseccomp-devel
-    python36 python3-virtualenv python3-pip nmap-ncat # for qa, macro_benchmark and stress tests
-    #
-    # IMPORTANT: python3-yaml does NOT exist on CentOS
-    # Install it manually using `pip3 install PyYAML`
-    #
-    PyYAML # Package name here does not correspond to the yum package!
-    libcurl-devel # mg-requests
-    rpm-build rpmlint # for RPM package building
-    doxygen graphviz # source documentation generators
-    which mono-complete dotnet-sdk-3.1 nodejs golang zip unzip java-11-openjdk-devel # for driver tests
-    sbcl # for custom Lisp C++ preprocessing
-    autoconf # for jemalloc code generation
-    libtool  # for protobuf code generation
-)
-
-list() {
-    echo "$1"
-}
-
-check() {
-    local missing=""
-    for pkg in $1; do
-        if [ "$pkg" == "PyYAML" ]; then
-            if ! python3 -c "import yaml" >/dev/null 2>/dev/null; then
-                missing="$pkg $missing"
-            fi
-            continue
-        fi
-        if ! yum list installed "$pkg" >/dev/null 2>/dev/null; then
-            missing="$pkg $missing"
-        fi
-    done
-    if [ "$missing" != "" ]; then
-        echo "MISSING PACKAGES: $missing"
-        exit 1
-    fi
-}
-
-install() {
-    cd "$DIR"
-    if [ "$EUID" -ne 0 ]; then
-        echo "Please run as root."
-        exit 1
-    fi
-    # If GitHub Actions runner is installed, append LANG to the environment.
-    # Python related tests doesn't work the LANG export.
-    if [ -d "/home/gh/actions-runner" ]; then
-        echo "LANG=en_US.utf8" >> /home/gh/actions-runner/.env
-    else
-        echo "NOTE: export LANG=en_US.utf8"
-    fi
-    dnf install -y epel-release
-    dnf install -y 'dnf-command(config-manager)'
-    dnf config-manager --set-enabled powertools # Required to install texinfo.
-    dnf update -y
-    dnf install -y wget git python36 python3-pip
-    for pkg in $1; do
-        if [ "$pkg" == libipt ]; then
-            if ! dnf list installed libipt >/dev/null 2>/dev/null; then
-                dnf install -y http://repo.okay.com.mx/centos/8/x86_64/release/libipt-1.6.1-8.el8.x86_64.rpm
-            fi
-            continue
-        fi
-        if [ "$pkg" == libipt-devel ]; then
-            if ! yum list installed libipt-devel >/dev/null 2>/dev/null; then
-                dnf install -y http://repo.okay.com.mx/centos/8/x86_64/release/libipt-devel-1.6.1-8.el8.x86_64.rpm
-            fi
-            continue
-        fi
-        # Install GDB dependencies not present in the standard repos.
-        # https://bugs.centos.org/view.php?id=17068
-        # https://centos.pkgs.org
-	    # Since 2020, there is Babeltrace2 (https://babeltrace.org). Not used
-        # within GDB yet (an assumption).
-        # http://mirror.centos.org/centos/8/PowerTools/x86_64/os/Packages/libbabeltrace-devel-1.5.4-3.el8.x86_64.rpm not working
-        if [ "$pkg" == libbabeltrace-devel ]; then
-            if ! dnf list installed libbabeltrace-devel >/dev/null 2>/dev/null; then
-                dnf install -y https://rpmfind.net/linux/centos/8-stream/PowerTools/x86_64/os/Packages/libbabeltrace-devel-1.5.4-3.el8.x86_64.rpm
-            fi
-            continue
-        fi
-        if [ "$pkg" == sbcl ]; then
-            if ! dnf list installed cl-asdf >/dev/null 2>/dev/null; then
-                dnf install -y 	https://pkgs.dyn.su/el8/base/x86_64/cl-asdf-20101028-18.el8.noarch.rpm
-            fi
-            if ! dnf list installed common-lisp-controller >/dev/null 2>/dev/null; then
-                dnf install -y https://pkgs.dyn.su/el8/base/x86_64/common-lisp-controller-7.4-20.el8.noarch.rpm
-            fi
-            if ! dnf list installed sbcl >/dev/null 2>/dev/null; then
-                dnf install -y https://pkgs.dyn.su/el8/base/x86_64/sbcl-2.0.1-4.el8.x86_64.rpm
-            fi
-            continue
-        fi
-        if [ "$pkg" == dotnet-sdk-3.1 ]; then
-            if ! dnf list installed dotnet-sdk-3.1 >/dev/null 2>/dev/null; then
-                wget -nv https://packages.microsoft.com/config/centos/8/packages-microsoft-prod.rpm -O packages-microsoft-prod.rpm
-                rpm -Uvh https://packages.microsoft.com/config/centos/8/packages-microsoft-prod.rpm
-                dnf update -y
-                dnf install -y dotnet-sdk-3.1
-            fi
-            continue
-        fi
-        if [ "$pkg" == PyYAML ]; then
-            if [ -z ${SUDO_USER+x} ]; then # Running as root (e.g. Docker).
-                pip3 install --user PyYAML
-            else # Running using sudo.
-                sudo -H -u "$SUDO_USER" bash -c "pip3 install --user PyYAML"
-            fi
-            continue
-        fi
-        dnf install -y "$pkg"
-    done
-}
-
-deps=$2"[*]"
-"$1" "${!deps}"

environment/os/centos-9.sh

@@ -6,14 +6,12 @@ DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
 source "$DIR/../util.sh"
 
 TOOLCHAIN_BUILD_DEPS=(
-    coreutils gcc gcc-c++ make # generic build tools
+    coreutils-common gcc gcc-c++ make # generic build tools
     wget # used for archive download
     gnupg2 # used for archive signature verification
     tar gzip bzip2 xz unzip # used for archive unpacking
     zlib-devel # zlib library used for all builds
-    expat-devel xz-devel python3-devel texinfo # for gdb
-    libcurl-devel # for cmake
-    curl # snappy
+    expat-devel xz-devel python3-devel texinfo libbabeltrace-devel # for gdb
     readline-devel # for cmake and llvm
     libffi-devel libxml2-devel # for llvm
     libedit-devel pcre-devel automake bison # for swig
@@ -21,6 +19,9 @@ TOOLCHAIN_BUILD_DEPS=(
     openssl-devel
     gmp-devel
     gperf
+    diffutils
+    libipt libipt-devel # intel
+    patch
 )
 
 TOOLCHAIN_RUN_DEPS=(
@@ -31,18 +32,19 @@ TOOLCHAIN_RUN_DEPS=(
     readline # for cmake and llvm
     libffi libxml2 # for llvm
     openssl-devel
+    perl # for openssl
 )
 
 MEMGRAPH_BUILD_DEPS=(
     git # source code control
     make pkgconf-pkg-config # build system
-    curl wget # for downloading libs
+    wget # for downloading libs
     libuuid-devel java-11-openjdk # required by antlr
     readline-devel # for memgraph console
     python3-devel # for query modules
     openssl-devel
     libseccomp-devel
-    python3 python3-virtualenv python3-pip nmap-ncat # for qa, macro_benchmark and stress tests
+    python3 python3-pip python3-virtualenv nmap-ncat # for qa, macro_benchmark and stress tests
     #
     # IMPORTANT: python3-yaml does NOT exist on CentOS
     # Install it manually using `pip3 install PyYAML`
@@ -73,12 +75,6 @@ check() {
         if [ "$pkg" == "python3-virtualenv" ]; then
             continue
         fi
-        if [ "$pkg" == sbcl ]; then
-            if ! sbcl --version &> /dev/null; then
-	        missing="$pkg $missing"
-            fi
-            continue
-        fi
         if ! yum list installed "$pkg" >/dev/null 2>/dev/null; then
             missing="$pkg $missing"
         fi
@@ -105,13 +101,37 @@ install() {
     yum update -y
     yum install -y wget git python3 python3-pip
     for pkg in $1; do
+        # Since there is no support for libipt-devel for CentOS 9 we install
+        # Fedoras version of same libs, they are the same version but released
+        # for different OS
+        # TODO Update when libipt-devel releases for CentOS 9
+        if [ "$pkg" == libipt ]; then
+            if ! dnf list installed libipt >/dev/null 2>/dev/null; then
+                dnf install -y http://repo.okay.com.mx/centos/8/x86_64/release/libipt-1.6.1-8.el8.x86_64.rpm
+            fi
+            continue
+        fi
+        if [ "$pkg" == libipt-devel ]; then
+            if ! dnf list installed libipt-devel >/dev/null 2>/dev/null; then
+                dnf install -y http://repo.okay.com.mx/centos/8/x86_64/release/libipt-devel-1.6.1-8.el8.x86_64.rpm
+            fi
+            continue
+        fi
+        if [ "$pkg" == libbabeltrace-devel ]; then
+            if ! dnf list installed libbabeltrace-devel >/dev/null 2>/dev/null; then
+                dnf install -y http://mirror.stream.centos.org/9-stream/CRB/x86_64/os/Packages/libbabeltrace-devel-1.5.8-10.el9.x86_64.rpm
+            fi
+            continue
+        fi
         if [ "$pkg" == sbcl ]; then
-            if ! sbcl --version &> /dev/null; then
-	        curl -s https://altushost-swe.dl.sourceforge.net/project/sbcl/sbcl/1.4.2/sbcl-1.4.2-arm64-linux-binary.tar.bz2 -o /tmp/sbcl-arm64.tar.bz2
-		tar xvjf /tmp/sbcl-arm64.tar.bz2 -C /tmp
-		pushd /tmp/sbcl-1.4.2-arm64-linux
-		INSTALL_ROOT=/usr/local sh install.sh
-		popd
+            if ! dnf list installed cl-asdf >/dev/null 2>/dev/null; then
+                dnf install -y 	https://pkgs.dyn.su/el8/base/x86_64/cl-asdf-20101028-18.el8.noarch.rpm
+            fi
+            if ! dnf list installed common-lisp-controller >/dev/null 2>/dev/null; then
+                dnf install -y https://pkgs.dyn.su/el8/base/x86_64/common-lisp-controller-7.4-20.el8.noarch.rpm
+            fi
+            if ! dnf list installed sbcl >/dev/null 2>/dev/null; then
+                dnf install -y https://pkgs.dyn.su/el8/base/x86_64/sbcl-2.0.1-4.el8.x86_64.rpm
             fi
             continue
         fi
@@ -125,9 +145,11 @@ install() {
         fi
         if [ "$pkg" == python3-virtualenv ]; then
             if [ -z ${SUDO_USER+x} ]; then # Running as root (e.g. Docker).
-                pip3 install --user virtualenv
+                pip3 install virtualenv
+                pip3 install virtualenvwrapper
             else # Running using sudo.
-                sudo -H -u "$SUDO_USER" bash -c "pip3 install --user virtualenv"
+                sudo -H -u "$SUDO_USER" bash -c "pip3 install virtualenv"
+                sudo -H -u "$SUDO_USER" bash -c "pip3 install virtualenvwrapper"
             fi
             continue
         fi

environment/os/debian-11-arm.sh

@@ -87,15 +87,6 @@ EOF
     fi
     apt install -y wget
     for pkg in $1; do
-        if [ "$pkg" == dotnet-sdk-3.1 ]; then
-            if ! dpkg -s "$pkg" 2>/dev/null >/dev/null; then
-                wget -nv https://packages.microsoft.com/config/debian/10/packages-microsoft-prod.deb -O packages-microsoft-prod.deb
-                dpkg -i packages-microsoft-prod.deb
-                apt-get update
-                apt-get install -y apt-transport-https dotnet-sdk-3.1
-            fi
-            continue
-        fi
         apt install -y "$pkg"
     done
 }

environment/os/debian-11.sh

@@ -87,7 +87,7 @@ EOF
     fi
     apt install -y wget
     for pkg in $1; do
-        if [ "$pkg" == dotnet-sdk-3.1 ]; then
+        if [ "$pkg" == dotnet-sdk-3.1  ]; then
             if ! dpkg -s "$pkg" 2>/dev/null >/dev/null; then
                 wget -nv https://packages.microsoft.com/config/debian/10/packages-microsoft-prod.deb -O packages-microsoft-prod.deb
                 dpkg -i packages-microsoft-prod.deb

environment/os/ubuntu-22.04.sh

@@ -0,0 +1,93 @@
+#!/bin/bash
+
+set -Eeuo pipefail
+
+DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
+source "$DIR/../util.sh"
+
+TOOLCHAIN_BUILD_DEPS=(
+    coreutils gcc g++ build-essential make # generic build tools
+    wget # used for archive download
+    gnupg # used for archive signature verification
+    tar gzip bzip2 xz-utils unzip # used for archive unpacking
+    zlib1g-dev # zlib library used for all builds
+    libexpat1-dev libipt-dev libbabeltrace-dev liblzma-dev python3-dev texinfo # for gdb
+    libcurl4-openssl-dev # for cmake
+    libreadline-dev # for cmake and llvm
+    libffi-dev libxml2-dev # for llvm
+    curl # snappy
+    file
+    git # for thrift
+    libgmp-dev # for gdb
+    gperf # for proxygen
+    libssl-dev
+    libedit-dev libpcre3-dev automake bison # for swig
+)
+
+TOOLCHAIN_RUN_DEPS=(
+    make # generic build tools
+    tar gzip bzip2 xz-utils # used for archive unpacking
+    zlib1g # zlib library used for all builds
+    libexpat1 libipt2 libbabeltrace1 liblzma5 python3 # for gdb
+    libcurl4 # for cmake
+    libreadline8 # for cmake and llvm
+    libffi7 libxml2 # for llvm
+    libssl-dev # for libevent
+)
+
+MEMGRAPH_BUILD_DEPS=(
+    git # source code control
+    make pkg-config # build system
+    curl wget # for downloading libs
+    uuid-dev default-jre-headless # required by antlr
+    libreadline-dev # for memgraph console
+    libpython3-dev python3-dev # for query modules
+    libssl-dev
+    libseccomp-dev
+    netcat # tests are using nc to wait for memgraph
+    python3 python3-virtualenv python3-pip # for qa, macro_benchmark and stress tests
+    python3-yaml # for the configuration generator
+    libcurl4-openssl-dev # mg-requests
+    sbcl # for custom Lisp C++ preprocessing
+    doxygen graphviz # source documentation generators
+    mono-runtime mono-mcs zip unzip default-jdk-headless # for driver tests
+    dotnet-sdk-6.0 golang nodejs npm
+    autoconf # for jemalloc code generation
+    libtool  # for protobuf code generation
+)
+
+list() {
+    echo "$1"
+}
+
+check() {
+    check_all_dpkg "$1"
+}
+
+install() {
+    cd "$DIR"
+    apt update
+    # If GitHub Actions runner is installed, append LANG to the environment.
+    # Python related tests doesn't work the LANG export.
+    if [ -d "/home/gh/actions-runner" ]; then
+        echo "LANG=en_US.utf8" >> /home/gh/actions-runner/.env
+    else
+        echo "NOTE: export LANG=en_US.utf8"
+    fi
+    apt install -y wget
+    for pkg in $1; do
+        if [ "$pkg" == dotnet-sdk-6.0 ]; then
+            if ! dpkg -s dotnet-sdk-6.0 2>/dev/null >/dev/null; then
+                wget -nv https://packages.microsoft.com/config/ubuntu/22.04/packages-microsoft-prod.deb -O packages-microsoft-prod.deb
+                dpkg -i packages-microsoft-prod.deb
+                apt-get update
+                apt-get install -y apt-transport-https dotnet-sdk-6.0
+            fi
+            continue
+        fi
+        apt install -y "$pkg"
+    done
+}
+
+deps=$2"[*]"
+"$1" "${!deps}"

environment/toolchain/v4.sh

@@ -675,7 +675,7 @@ PROXYGEN_SHA256=5360a8ccdfb2f5a6c7b3eed331ec7ab0e2c792d579c6fff499c85c516c11fe14
 SNAPPY_SHA256=75c1fbb3d618dd3a0483bff0e26d0a92b495bbe5059c8b4f1c962b478b6e06e7
 SNAPPY_VERSION=1.1.9
 XZ_VERSION=5.2.5 # for LZMA
-ZLIB_VERSION=1.2.11
+ZLIB_VERSION=1.2.12
 ZSTD_VERSION=1.5.0
 WANGLE_SHA256=1002e9c32b6f4837f6a760016e3b3e22f3509880ef3eaad191c80dc92655f23f
 
@@ -1178,12 +1178,21 @@ popd
 
 # create toolchain archive
 if [ ! -f $NAME-binaries-$DISTRO.tar.gz ]; then
-    DISTRO_FULL_NAME=$DISTRO
-    if [ "$for_arm" = true ]; then
-        DISTRO_FULL_NAME="$DISTRO_FULL_NAME-aarch64"
+    DISTRO_FULL_NAME=${DISTRO}
+    if [[ "${DISTRO}" == centos* ]]; then
+        if [[ "$for_arm" = "true" ]]; then
+            DISTRO_FULL_NAME="$DISTRO_FULL_NAME-aarch64"
+        else
+            DISTRO_FULL_NAME="$DISTRO_FULL_NAME-x86_64"
+        fi
     else
-        DISTRO_FULL_NAME="$DISTRO_FULL_NAME-x86_64"
+        if [[ "$for_arm" = "true" ]]; then
+            DISTRO_FULL_NAME="$DISTRO_FULL_NAME-arm64"
+        else
+            DISTRO_FULL_NAME="$DISTRO_FULL_NAME-amd64"
+        fi
     fi
+
     tar --owner=root --group=root -cpvzf $NAME-binaries-$DISTRO_FULL_NAME.tar.gz -C /opt $NAME
 fi
 

init

@@ -24,7 +24,7 @@ function setup_virtualenv () {
     fi
 
     # create new virtualenv
-    virtualenv -p python3 ve3 || exit 1
+    python3 -m virtualenv -p python3 ve3 || exit 1
     source ve3/bin/activate
     pip --timeout 1000 install -r requirements.txt || exit 1
     deactivate
@@ -66,7 +66,7 @@ fi
 
 DISTRO=$(operating_system)
 ARCHITECTURE=$(architecture)
-if [ "${ARCHITECTURE}" = "arm64" ]; then
+if [ "${ARCHITECTURE}" = "arm64" ] || [ "${ARCHITECTURE}" = "aarch64" ]; then
   OS_SCRIPT=$DIR/environment/os/$DISTRO-arm.sh
 else
   OS_SCRIPT=$DIR/environment/os/$DISTRO.sh
@@ -135,3 +135,10 @@ for hook in $(find $DIR/.githooks -type f -printf "%f\n"); do
   ln -s -f "$DIR/.githooks/$hook" "$DIR/.git/hooks/$hook"
   echo "Added $hook hook"
 done;
+
+# Install precommit hook
+python3 -m pip install pre-commit
+python3 -m pre_commit install
+
+# Link `include/mgp.py` with `release/mgp/mgp.py`
+ln -v -f include/mgp.py release/mgp/mgp.py

libs/.gitignore

@@ -4,5 +4,4 @@
 !cleanup.sh
 !CMakeLists.txt
 !__main.cpp
-!jemalloc.cmake
 !pulsar.patch

libs/jemalloc.cmake

@@ -1,55 +0,0 @@
-set(JEMALLOC_DIR "${LIB_DIR}/jemalloc")
-
-set(JEMALLOC_SRCS
-    ${JEMALLOC_DIR}/src/arena.c
-    ${JEMALLOC_DIR}/src/background_thread.c
-    ${JEMALLOC_DIR}/src/base.c
-    ${JEMALLOC_DIR}/src/bin.c
-    ${JEMALLOC_DIR}/src/bitmap.c
-    ${JEMALLOC_DIR}/src/ckh.c
-    ${JEMALLOC_DIR}/src/ctl.c
-    ${JEMALLOC_DIR}/src/div.c
-    ${JEMALLOC_DIR}/src/extent.c
-    ${JEMALLOC_DIR}/src/extent_dss.c
-    ${JEMALLOC_DIR}/src/extent_mmap.c
-    ${JEMALLOC_DIR}/src/hash.c
-    ${JEMALLOC_DIR}/src/hook.c
-    ${JEMALLOC_DIR}/src/jemalloc.c
-    ${JEMALLOC_DIR}/src/large.c
-    ${JEMALLOC_DIR}/src/log.c
-    ${JEMALLOC_DIR}/src/malloc_io.c
-    ${JEMALLOC_DIR}/src/mutex.c
-    ${JEMALLOC_DIR}/src/mutex_pool.c
-    ${JEMALLOC_DIR}/src/nstime.c
-    ${JEMALLOC_DIR}/src/pages.c
-    ${JEMALLOC_DIR}/src/prng.c
-    ${JEMALLOC_DIR}/src/prof.c
-    ${JEMALLOC_DIR}/src/rtree.c
-    ${JEMALLOC_DIR}/src/sc.c
-    ${JEMALLOC_DIR}/src/stats.c
-    ${JEMALLOC_DIR}/src/sz.c
-    ${JEMALLOC_DIR}/src/tcache.c
-    ${JEMALLOC_DIR}/src/test_hooks.c
-    ${JEMALLOC_DIR}/src/ticker.c
-    ${JEMALLOC_DIR}/src/tsd.c
-    ${JEMALLOC_DIR}/src/witness.c
-    ${JEMALLOC_DIR}/src/safety_check.c
-)
-
-add_library(jemalloc ${JEMALLOC_SRCS})
-target_include_directories(jemalloc PUBLIC "${JEMALLOC_DIR}/include")
-
-find_package(Threads REQUIRED)
-target_link_libraries(jemalloc PUBLIC Threads::Threads)
-
-target_compile_definitions(jemalloc PRIVATE -DJEMALLOC_NO_PRIVATE_NAMESPACE)
-
-if (CMAKE_BUILD_TYPE STREQUAL "DEBUG")
-    target_compile_definitions(jemalloc PRIVATE -DJEMALLOC_DEBUG=1 -DJEMALLOC_PROF=1)
-endif()
-
-target_compile_options(jemalloc PRIVATE -Wno-redundant-decls)
-# for RTLD_NEXT
-target_compile_definitions(jemalloc PRIVATE _GNU_SOURCE)
-
-set_property(TARGET jemalloc APPEND PROPERTY INTERFACE_COMPILE_DEFINITIONS USE_JEMALLOC=1)

release/CMakeLists.txt

@@ -41,7 +41,7 @@ set(CPACK_DEBIAN_PACKAGE_DESCRIPTION "${CPACK_PACKAGE_DESCRIPTION_SUMMARY}
  applications driver by real-time connected data.")
 # Add `openssl` package to dependencies list. Used to generate SSL certificates.
 # We also depend on `python3` because we embed it in Memgraph.
-set(CPACK_DEBIAN_PACKAGE_DEPENDS "openssl (>= 1.1.0), python3 (>= 3.5.0)")
+set(CPACK_DEBIAN_PACKAGE_DEPENDS "openssl (>= 1.1.0), python3 (>= 3.5.0), libstdc++6")
 
 # Setting arhitecture extension for rpm packages
 set(MG_ARCH_EXTENSION_RPM "noarch")
@@ -67,7 +67,7 @@ It aims to deliver developers the speed, simplicity and scale required to build
 the next generation of applications driver by real-time connected data.")
 # Add `openssl` package to dependencies list. Used to generate SSL certificates.
 # We also depend on `python3` because we embed it in Memgraph.
-set(CPACK_RPM_PACKAGE_REQUIRES "openssl >= 1.0.0, curl >= 7.29.0, python3 >= 3.5.0")
+set(CPACK_RPM_PACKAGE_REQUIRES "openssl >= 1.0.0, curl >= 7.29.0, python3 >= 3.5.0, libstdc >= 6, logrotate")
 
 # All variables must be set before including.
 include(CPack)

release/docker/memgraph_rpm.dockerfile

@@ -1,31 +0,0 @@
-FROM dokken/centos-stream-9
-# NOTE: If you change the base distro update release/package as well.
-
-ARG release
-
-RUN yum update && yum install -y \
-    openssl libcurl libseccomp python3 python3-pip \
-    --nobest --allowerasing \
-  && rm -rf /tmp/* \
-  && yum clean all
-
-RUN pip3 install networkx==2.4 numpy==1.21.4 scipy==1.7.3
-
-COPY ${release} /
-
-# Install memgraph package
-RUN rpm -i ${release}
-
-# Memgraph listens for Bolt Protocol on this port by default.
-EXPOSE 7687
-# Snapshots and logging volumes
-VOLUME /var/log/memgraph
-VOLUME /var/lib/memgraph
-# Configuration volume
-VOLUME /etc/memgraph
-
-USER memgraph
-WORKDIR /usr/lib/memgraph
-
-ENTRYPOINT ["/usr/lib/memgraph/memgraph"]
-CMD [""]

release/mgp/.gitignore

@@ -0,0 +1,3 @@
+.venv
+dist
+mgp.py

release/mgp/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

release/mgp/README.md

@@ -0,0 +1,4 @@
+# mgp
+
+PyPi package used for type hinting when creating MAGE modules. The get started
+using MAGE repository checkout the repository here: https://github.com/memgraph/mage.

release/mgp/_mgp.py

@@ -0,0 +1,255 @@
+from typing import Any
+
+
+class MgpIterable:
+    def get() -> Any:
+        pass
+
+    def next() -> Any:
+        pass
+
+
+class Vertex:
+    def is_valid() -> bool:  # type: ignore
+        pass
+
+    def underlying_graph_is_mutable() -> bool:  # type: ignore
+        pass
+
+    def iter_properties() -> MgpIterable:  # type: ignore
+        pass
+
+    def get_property(self, property_name: str) -> "Property":  # type: ignore
+        pass
+
+    def set_property(self, property_name: str, value: Any) -> "Property":  # type: ignore
+        pass
+
+    def get_id() -> "VertexId":  # type: ignore
+        pass
+
+    def label_at(self, index: int) -> "Label":  # type: ignore
+        pass
+
+    def labels_count() -> int:  # type: ignore
+        pass
+
+    def add_label(self, label: Any):
+        pass
+
+    def remove_label(self, label: Any):
+        pass
+
+    def iter_in_edges() -> MgpIterable:  # type: ignore
+        pass
+
+    def iter_out_edges() -> MgpIterable:  # type: ignore
+        pass
+
+
+class Edge:
+    def is_valid() -> bool:  # type: ignore
+        pass
+
+    def underlying_graph_is_mutable() -> bool:  # type: ignore
+        pass
+
+    def iter_properties() -> MgpIterable:  # type: ignore
+        pass
+
+    def get_property(self, property_name: str) -> "Property":  # type: ignore
+        pass
+
+    def set_property(self, property_name: str, valuse: Any) -> "Property":  # type: ignore
+        pass
+
+    def get_type_name() -> str:  # type: ignore
+        pass
+
+    def get_id() -> "EdgeId":  # type: ignore
+        pass
+
+    def from_vertex() -> Vertex:  # type: ignore
+        pass
+
+    def to_vertex() -> Vertex:  # type: ignore
+        pass
+
+
+class Path:
+    def is_valid() -> bool:  # type: ignore
+        pass
+
+    @staticmethod
+    def make_with_start(vertex: Vertex) -> "Path":  # type: ignore
+        pass
+
+
+class Graph:
+    def is_valid() -> bool:  # type: ignore
+        pass
+
+
+class CypherType:
+    pass
+
+
+class Message:
+    def is_valid() -> bool:  # type: ignore
+        pass
+
+    def source_type() -> str:  # type: ignore
+        pass
+
+    def topic_name() -> str:  # type: ignore
+        pass
+
+    def key() -> bytes:  # type: ignore
+        pass
+
+    def timestamp() -> int:  # type: ignore
+        pass
+
+    def offset() -> int:  # type: ignore
+        pass
+
+    def payload() -> bytes:  # type: ignore
+        pass
+
+
+class Messages:
+    def is_valid() -> bool:  # type: ignore
+        pass
+
+    def message_at(self, id: int) -> Message:  # type: ignore
+        pass
+
+    def total_messages() -> int:  # type: ignore
+        pass
+
+
+class UnknownError(Exception):
+    pass
+
+
+class UnableToAllocateError(Exception):
+    pass
+
+
+class InsufficientBufferError(Exception):
+    pass
+
+
+class OutOfRangeError(Exception):
+    pass
+
+
+class LogicErrorError(Exception):
+
+    pass
+
+
+class DeletedObjectError(Exception):
+    pass
+
+
+class InvalidArgumentError(Exception):
+    pass
+
+
+class KeyAlreadyExistsError(Exception):
+    pass
+
+
+class ImmutableObjectError(Exception):
+    pass
+
+
+class ValueConversionError(Exception):
+    pass
+
+
+class SerializationError(Exception):
+    pass
+
+
+def type_nullable(elem: Any):
+    pass
+
+
+def type_list(elem: Any):
+    pass
+
+
+def type_bool():
+    pass
+
+
+def type_string():
+    pass
+
+
+def type_int():
+    pass
+
+
+def type_float():
+    pass
+
+
+def type_number():
+    pass
+
+
+def type_map():
+    pass
+
+
+def type_node():
+    pass
+
+
+def type_relationship():
+    pass
+
+
+def type_path():
+    pass
+
+
+def type_date():
+    pass
+
+
+def type_local_time():
+    pass
+
+
+def type_local_date_time():
+    pass
+
+
+def type_duration():
+    pass
+
+
+def type_any():
+    pass
+
+
+class _MODULE:
+    @staticmethod
+    def add_read_procedure(wrapper):
+        pass
+
+    @staticmethod
+    def add_write_procedure(wrapper):
+        pass
+
+    @staticmethod
+    def add_transformation(wrapper):
+        pass
+
+    @staticmethod
+    def add_function(wrapper):
+        pass

release/mgp/instructions.md

@@ -0,0 +1,22 @@
+# How to publish new versions
+## Prerequisites
+1. Installed poetry
+```
+pip install poetry
+```
+2. Set up [API tokens](https://pypi.org/help/#apitoken)
+3. Be a collaborator on [pypi](https://pypi.org/project/mgp/)
+
+## Making changes
+1. Make changes to the package
+2. Bump version in `pyproject.tml`
+3. `poetry build`
+4. `poetry publish`
+
+## Why is this not automatized?
+
+Because someone always has to manually bump up the version in `pyproject.toml`
+
+## Why does `_mgp.py` exists?
+Because we are mocking here all the types that are created by Memgraph
+in order to fix typing errors in `mgp.py`.

release/mgp/pyproject.toml

@@ -0,0 +1,23 @@
+[tool.poetry]
+name = "mgp"
+version = "1.0.0"
+description = "Memgraph's module for developing MAGE modules. Used only for type hinting!"
+authors = [
+    "MasterMedo <mislav.vuletic@gmail.com>",
+    "jbajic <jure.bajic@memgraph.io>",
+    "katarinasupe <katarina.supe@memgraph.io>",
+    "antejavor <ante.javor@memgraph.io>",
+    "antaljanosbenjamin <benjamin.antal@memgraph.io>",
+]
+license = "Apache-2.0"
+readme = "README.md"
+include = ["mgp.py", "_mgp.py"]
+
+[tool.poetry.dependencies]
+python = "^3.7"
+
+[tool.poetry.dev-dependencies]
+
+[build-system]
+requires = ["poetry-core>=1.0.0"]
+build-backend = "poetry.core.masonry.api"

release/package/centos-7/Dockerfile

@@ -7,8 +7,8 @@ RUN yum -y update \
 # Do NOT be smart here and clean the cache because the container is used in the
 # stateful context.
 
-RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-centos-7.tar.gz \
-    -O ${TOOLCHAIN_VERSION}-binaries-centos-7.tar.gz \
-    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-centos-7.tar.gz -C /opt
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-centos-7-x86_64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-centos-7-x86_64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-centos-7-x86_64.tar.gz -C /opt
 
 ENTRYPOINT ["sleep", "infinity"]

release/package/centos-8/Dockerfile

@@ -1,14 +0,0 @@
-FROM centos:8
-
-ARG TOOLCHAIN_VERSION
-
-RUN dnf -y update \
-    && dnf install -y wget git
-# Do NOT be smart here and clean the cache because the container is used in the
-# stateful context.
-
-RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-centos-8.tar.gz \
-    -O ${TOOLCHAIN_VERSION}-binaries-centos-8.tar.gz \
-    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-centos-8.tar.gz -C /opt
-
-ENTRYPOINT ["sleep", "infinity"]

release/package/centos-9/Dockerfile

@@ -0,0 +1,14 @@
+FROM quay.io/centos/centos:stream9
+
+ARG TOOLCHAIN_VERSION
+
+RUN yum -y update \
+    && yum install -y wget git
+# Do NOT be smart here and clean the cache because the container is used in the
+# stateful context.
+
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-centos-9-x86_64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-centos-9-x86_64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-centos-9-x86_64.tar.gz -C /opt
+
+ENTRYPOINT ["sleep", "infinity"]

release/package/debian-10/Dockerfile

@@ -10,8 +10,8 @@ RUN apt update && apt install -y \
 # Do NOT be smart here and clean the cache because the container is used in the
 # stateful context.
 
-RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-debian-10.tar.gz \
-    -O ${TOOLCHAIN_VERSION}-binaries-debian-10.tar.gz \
-    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-debian-10.tar.gz -C /opt
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-debian-10-amd64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-debian-10-amd64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-debian-10-amd64.tar.gz -C /opt
 
 ENTRYPOINT ["sleep", "infinity"]

release/package/debian-11-arm/Dockerfile

@@ -0,0 +1,17 @@
+FROM debian:11
+
+ARG TOOLCHAIN_VERSION
+
+# Stops tzdata interactive configuration.
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN apt update && apt install -y \
+    ca-certificates wget git
+# Do NOT be smart here and clean the cache because the container is used in the
+# stateful context.
+
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-debian-11-arm64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-debian-11-arm64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-debian-11-arm64.tar.gz -C /opt
+
+ENTRYPOINT ["sleep", "infinity"]

release/package/debian-11/Dockerfile

@@ -10,8 +10,8 @@ RUN apt update && apt install -y \
 # Do NOT be smart here and clean the cache because the container is used in the
 # stateful context.
 
-RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-debian-11.tar.gz \
-    -O ${TOOLCHAIN_VERSION}-binaries-debian-11.tar.gz \
-    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-debian-11.tar.gz -C /opt
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-debian-11-amd64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-debian-11-amd64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-debian-11-amd64.tar.gz -C /opt
 
 ENTRYPOINT ["sleep", "infinity"]

release/package/docker-compose.yml

@@ -4,10 +4,10 @@ services:
     build:
       context: centos-7
     container_name: "mgbuild_centos-7"
-  mgbuild_centos-8:
+  mgbuild_centos-9:
     build:
-      context: centos-8
-    container_name: "mgbuild_centos-8"
+      context: centos-9
+    container_name: "mgbuild_centos-9"
   mgbuild_debian-10:
     build:
       context: debian-10
@@ -24,3 +24,7 @@ services:
     build:
       context: ubuntu-20.04
     container_name: "mgbuild_ubuntu-20.04"
+  mgbuild_ubuntu-22.04:
+    build:
+      context: ubuntu-22.04
+    container_name: "mgbuild_ubuntu-22.04"

release/package/run.sh

@@ -3,7 +3,7 @@
 set -Eeuo pipefail
 
 SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
-SUPPORTED_OS=(centos-7 centos-8 debian-10 debian-11 ubuntu-18.04 ubuntu-20.04)
+SUPPORTED_OS=(centos-7 centos-9 debian-10 debian-11 ubuntu-18.04 ubuntu-20.04 ubuntu-22.04 debian-11-arm)
 PROJECT_ROOT="$SCRIPT_DIR/../.."
 TOOLCHAIN_VERSION="toolchain-v4"
 ACTIVATE_TOOLCHAIN="source /opt/${TOOLCHAIN_VERSION}/activate"
@@ -67,14 +67,18 @@ make_package () {
     # environment/os/{os}.sh does not come within the toolchain package. When
     # migrating to the next version of toolchain do that, and remove the
     # TOOLCHAIN_RUN_DEPS installation from here.
-    echo "Installing dependencies..."
+    echo "Installing dependencies using '/memgraph/environment/os/$os.sh' script..."
     docker exec "$build_container" bash -c "/memgraph/environment/os/$os.sh install TOOLCHAIN_RUN_DEPS"
     docker exec "$build_container" bash -c "/memgraph/environment/os/$os.sh install MEMGRAPH_BUILD_DEPS"
 
     echo "Building targeted package..."
     docker exec "$build_container" bash -c "cd /memgraph && $ACTIVATE_TOOLCHAIN && ./init"
     docker exec "$build_container" bash -c "cd $container_build_dir && rm -rf ./*"
-    docker exec "$build_container" bash -c "cd $container_build_dir && $ACTIVATE_TOOLCHAIN && cmake -DCMAKE_BUILD_TYPE=release $telemetry_id_override_flag .."
+    if [[ "$os" == "debian-11-arm" ]]; then
+        docker exec "$build_container" bash -c "cd $container_build_dir && $ACTIVATE_TOOLCHAIN && cmake -DCMAKE_BUILD_TYPE=release -DMG_ARCH="ARM64" $telemetry_id_override_flag .."
+    else
+        docker exec "$build_container" bash -c "cd $container_build_dir && $ACTIVATE_TOOLCHAIN && cmake -DCMAKE_BUILD_TYPE=release $telemetry_id_override_flag .."
+    fi
     # ' is used instead of " because we need to run make within the allowed
     # container resources.
     # shellcheck disable=SC2016

release/package/ubuntu-18.04/Dockerfile

@@ -10,8 +10,8 @@ RUN apt update && apt install -y \
 # Do NOT be smart here and clean the cache because the container is used in the
 # stateful context.
 
-RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-ubuntu-18.04.tar.gz \
-    -O ${TOOLCHAIN_VERSION}-binaries-ubuntu-18.04.tar.gz \
-    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-ubuntu-18.04.tar.gz -C /opt
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-ubuntu-18.04-amd64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-ubuntu-18.04-amd64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-ubuntu-18.04-amd64.tar.gz -C /opt
 
 ENTRYPOINT ["sleep", "infinity"]

release/package/ubuntu-20.04/Dockerfile

@@ -10,8 +10,8 @@ RUN apt update && apt install -y \
 # Do NOT be smart here and clean the cache because the container is used in the
 # stateful context.
 
-RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-ubuntu-20.04.tar.gz \
-    -O ${TOOLCHAIN_VERSION}-binaries-ubuntu-20.04.tar.gz \
-    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-ubuntu-20.04.tar.gz -C /opt
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-ubuntu-20.04-amd64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-ubuntu-20.04-amd64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-ubuntu-20.04-amd64.tar.gz -C /opt
 
 ENTRYPOINT ["sleep", "infinity"]

release/package/ubuntu-22.04/Dockerfile

@@ -0,0 +1,17 @@
+FROM ubuntu:22.04
+
+ARG TOOLCHAIN_VERSION
+
+# Stops tzdata interactive configuration.
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN apt update && apt install -y \
+    ca-certificates wget git
+# Do NOT be smart here and clean the cache because the container is used in the
+# stateful context.
+
+RUN wget -q https://s3-eu-west-1.amazonaws.com/deps.memgraph.io/${TOOLCHAIN_VERSION}/${TOOLCHAIN_VERSION}-binaries-ubuntu-22.04-amd64.tar.gz \
+    -O ${TOOLCHAIN_VERSION}-binaries-ubuntu-22.04-amd64.tar.gz \
+    && tar xzvf ${TOOLCHAIN_VERSION}-binaries-ubuntu-22.04-amd64.tar.gz -C /opt
+
+ENTRYPOINT ["sleep", "infinity"]

src/communication/v2/session.hpp

@@ -105,9 +105,16 @@ class WebsocketSession : public std::enable_shared_from_this<WebsocketSession<TS
     boost::asio::socket_base::keep_alive option(true);
 
     // Set a decorator to change the Server of the handshake
-    ws_.set_option(boost::beast::websocket::stream_base::decorator([](boost::beast::websocket::response_type &res) {
+    ws_.set_option(boost::beast::websocket::stream_base::decorator([&req](boost::beast::websocket::response_type &res) {
       res.set(boost::beast::http::field::server, std::string("Memgraph Bolt WS"));
-      res.set(boost::beast::http::field::sec_websocket_protocol, "binary");
+
+      // We need to do this to support WASM clients, which explicitly send this flag
+      // in their upgrade request
+      // Neo4j client breaks when this flag is sent
+      if (const auto secondary_protocol = req.base().find(boost::beast::http::field::sec_websocket_protocol);
+          secondary_protocol != res.base().end() && secondary_protocol->value() == "binary") {
+        res.set(boost::beast::http::field::sec_websocket_protocol, "binary");
+      }
     }));
     ws_.binary(true);
 
@@ -162,7 +169,7 @@ class WebsocketSession : public std::enable_shared_from_this<WebsocketSession<TS
         boost::asio::bind_executor(strand_, std::bind_front(&WebsocketSession::OnRead, shared_from_this())));
   }
 
-  void OnRead(const boost::system::error_code &ec, [[maybe_unused]] const size_t bytes_transferred) {
+  void OnRead(const boost::system::error_code &ec, const size_t bytes_transferred) {
     // This indicates that the WebsocketSession was closed
     if (ec == boost::beast::websocket::error::closed) {
       return;
@@ -246,11 +253,7 @@ class Session final : public std::enable_shared_from_this<Session<TSession, TSes
   Session(Session &&) = delete;
   Session &operator=(const Session &) = delete;
   Session &operator=(Session &&) = delete;
-  ~Session() {
-    if (IsConnected()) {
-      spdlog::error("Session: Destructor called while execution is active");
-    }
-  }
+  ~Session() = default;
 
   bool Start() {
     if (execution_active_) {
@@ -326,7 +329,8 @@ class Session final : public std::enable_shared_from_this<Session<TSession, TSes
       socket.lowest_layer().non_blocking(false);
     });
     timeout_timer_.expires_at(boost::asio::steady_timer::time_point::max());
-    spdlog::info("Accepted a connection from {}:", service_name_, remote_endpoint_.address(), remote_endpoint_.port());
+    spdlog::info("Accepted a connection from {}: {}:{}", service_name_, remote_endpoint_.address(),
+                 remote_endpoint_.port());
   }
 
   void DoRead() {
@@ -400,7 +404,6 @@ class Session final : public std::enable_shared_from_this<Session<TSession, TSes
     if (ec == boost::asio::error::operation_aborted) {
       return;
     }
-    execution_active_ = false;
 
     if (ec == boost::asio::error::eof) {
       spdlog::info("Session closed by peer");

src/integrations/constants.hpp

@@ -17,6 +17,7 @@
 namespace memgraph::integrations {
 
 inline constexpr int64_t kDefaultCheckBatchLimit{1};
+inline constexpr int64_t kMinimumStartBatchLimit{1};
 inline constexpr std::chrono::milliseconds kDefaultCheckTimeout{30000};
 inline constexpr std::chrono::milliseconds kMinimumInterval{1};
 inline constexpr int64_t kMinimumSize{1};

src/integrations/kafka/consumer.cpp

@@ -74,6 +74,36 @@ utils::BasicResult<std::string, std::vector<Message>> GetBatch(RdKafka::KafkaCon
 
   return std::move(batch);
 }
+
+void CheckAndDestroyLastAssignmentIfNeeded(RdKafka::KafkaConsumer &consumer, const ConsumerInfo &info,
+                                           std::vector<RdKafka::TopicPartition *> &last_assignment) {
+  if (!last_assignment.empty()) {
+    if (const auto err = consumer.assign(last_assignment); err != RdKafka::ERR_NO_ERROR) {
+      throw ConsumerStartFailedException(info.consumer_name,
+                                         fmt::format("Couldn't restore commited offsets: '{}'", RdKafka::err2str(err)));
+    }
+    RdKafka::TopicPartition::destroy(last_assignment);
+  }
+}
+
+void TryToConsumeBatch(RdKafka::KafkaConsumer &consumer, const ConsumerInfo &info,
+                       const ConsumerFunction &consumer_function, const std::vector<Message> &batch) {
+  consumer_function(batch);
+  std::vector<RdKafka::TopicPartition *> partitions;
+  utils::OnScopeExit clear_partitions([&]() { RdKafka::TopicPartition::destroy(partitions); });
+
+  if (const auto err = consumer.assignment(partitions); err != RdKafka::ERR_NO_ERROR) {
+    throw ConsumerCommitFailedException(
+        info.consumer_name, fmt::format("Couldn't get assignment to commit offsets: {}", RdKafka::err2str(err)));
+  }
+  if (const auto err = consumer.position(partitions); err != RdKafka::ERR_NO_ERROR) {
+    throw ConsumerCommitFailedException(info.consumer_name,
+                                        fmt::format("Couldn't get offsets from librdkafka {}", RdKafka::err2str(err)));
+  }
+  if (const auto err = consumer.commitSync(partitions); err != RdKafka::ERR_NO_ERROR) {
+    throw ConsumerCommitFailedException(info.consumer_name, RdKafka::err2str(err));
+  }
+}
 }  // namespace
 
 Message::Message(std::unique_ptr<RdKafka::Message> &&message) : message_{std::move(message)} {
@@ -221,10 +251,21 @@ void Consumer::Start() {
   StartConsuming();
 }
 
-void Consumer::StartIfStopped() {
-  if (!is_running_) {
-    StartConsuming();
+void Consumer::StartWithLimit(const uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const {
+  if (is_running_) {
+    throw ConsumerRunningException(info_.consumer_name);
   }
+  if (limit_batches < kMinimumStartBatchLimit) {
+    throw ConsumerStartFailedException(
+        info_.consumer_name, fmt::format("Batch limit has to be greater than or equal to {}", kMinimumStartBatchLimit));
+  }
+  if (timeout.value_or(kMinimumInterval) < kMinimumInterval) {
+    throw ConsumerStartFailedException(
+        info_.consumer_name,
+        fmt::format("Timeout has to be greater than or equal to {} milliseconds", kMinimumInterval.count()));
+  }
+
+  StartConsumingWithLimit(limit_batches, timeout);
 }
 
 void Consumer::Stop() {
@@ -244,7 +285,7 @@ void Consumer::StopIfRunning() {
   }
 }
 
-void Consumer::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> limit_batches,
+void Consumer::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> limit_batches,
                      const ConsumerFunction &check_consumer_function) const {
   // NOLINTNEXTLINE (modernize-use-nullptr)
   if (timeout.value_or(kMinimumInterval) < kMinimumInterval) {
@@ -344,13 +385,7 @@ void Consumer::StartConsuming() {
 
   is_running_.store(true);
 
-  if (!last_assignment_.empty()) {
-    if (const auto err = consumer_->assign(last_assignment_); err != RdKafka::ERR_NO_ERROR) {
-      throw ConsumerStartFailedException(info_.consumer_name,
-                                         fmt::format("Couldn't restore commited offsets: '{}'", RdKafka::err2str(err)));
-    }
-    RdKafka::TopicPartition::destroy(last_assignment_);
-  }
+  CheckAndDestroyLastAssignmentIfNeeded(*consumer_, info_, last_assignment_);
 
   thread_ = std::thread([this] {
     static constexpr auto kMaxThreadNameSize = utils::GetMaxThreadNameSize();
@@ -361,33 +396,18 @@ void Consumer::StartConsuming() {
     while (is_running_) {
       auto maybe_batch = GetBatch(*consumer_, info_, is_running_);
       if (maybe_batch.HasError()) {
-        spdlog::warn("Error happened in consumer {} while fetching messages: {}!", info_.consumer_name,
-                     maybe_batch.GetError());
-        break;
+        throw ConsumerReadMessagesFailedException(info_.consumer_name, maybe_batch.GetError());
       }
       const auto &batch = maybe_batch.GetValue();
 
-      if (batch.empty()) continue;
+      if (batch.empty()) {
+        continue;
+      }
 
       spdlog::info("Kafka consumer {} is processing a batch", info_.consumer_name);
 
       try {
-        consumer_function_(batch);
-        std::vector<RdKafka::TopicPartition *> partitions;
-        utils::OnScopeExit clear_partitions([&]() { RdKafka::TopicPartition::destroy(partitions); });
-
-        if (const auto err = consumer_->assignment(partitions); err != RdKafka::ERR_NO_ERROR) {
-          throw ConsumerCheckFailedException(
-              info_.consumer_name, fmt::format("Couldn't get assignment to commit offsets: {}", RdKafka::err2str(err)));
-        }
-        if (const auto err = consumer_->position(partitions); err != RdKafka::ERR_NO_ERROR) {
-          throw ConsumerCheckFailedException(
-              info_.consumer_name, fmt::format("Couldn't get offsets from librdkafka {}", RdKafka::err2str(err)));
-        }
-        if (const auto err = consumer_->commitSync(partitions); err != RdKafka::ERR_NO_ERROR) {
-          spdlog::warn("Committing offset of consumer {} failed: {}", info_.consumer_name, RdKafka::err2str(err));
-          break;
-        }
+        TryToConsumeBatch(*consumer_, info_, consumer_function_, batch);
       } catch (const std::exception &e) {
         spdlog::warn("Error happened in consumer {} while processing a batch: {}!", info_.consumer_name, e.what());
         break;
@@ -398,6 +418,44 @@ void Consumer::StartConsuming() {
   });
 }
 
+void Consumer::StartConsumingWithLimit(uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const {
+  MG_ASSERT(!is_running_, "Cannot start already running consumer!");
+
+  if (is_running_.exchange(true)) {
+    throw ConsumerRunningException(info_.consumer_name);
+  }
+  utils::OnScopeExit restore_is_running([this] { is_running_.store(false); });
+
+  CheckAndDestroyLastAssignmentIfNeeded(*consumer_, info_, last_assignment_);
+
+  const auto timeout_to_use = timeout.value_or(kDefaultCheckTimeout);
+  const auto start = std::chrono::steady_clock::now();
+
+  for (uint64_t batch_count = 0; batch_count < limit_batches;) {
+    const auto now = std::chrono::steady_clock::now();
+    if (now - start >= timeout_to_use) {
+      throw ConsumerStartFailedException(info_.consumer_name, "Timeout reached");
+    }
+
+    const auto maybe_batch = GetBatch(*consumer_, info_, is_running_);
+    if (maybe_batch.HasError()) {
+      throw ConsumerReadMessagesFailedException(info_.consumer_name, maybe_batch.GetError());
+    }
+    const auto &batch = maybe_batch.GetValue();
+
+    if (batch.empty()) {
+      continue;
+    }
+    ++batch_count;
+
+    spdlog::info("Kafka consumer {} is processing a batch", info_.consumer_name);
+
+    TryToConsumeBatch(*consumer_, info_, consumer_function_, batch);
+
+    spdlog::info("Kafka consumer {} finished processing", info_.consumer_name);
+  }
+}
+
 void Consumer::StopConsuming() {
   is_running_.store(false);
   if (thread_.joinable()) thread_.join();

src/integrations/kafka/consumer.hpp

@@ -113,11 +113,19 @@ class Consumer final : public RdKafka::EventCb {
   /// This method will start a new thread which will poll all the topics for messages.
   ///
   /// @throws ConsumerRunningException if the consumer is already running
+  /// @throws ConsumerStartFailedException if the commited offsets cannot be restored
   void Start();
 
-  /// Starts consuming messages if it is not started already.
+  /// Starts consuming messages.
   ///
-  void StartIfStopped();
+  /// This method will start a new thread which will poll all the topics for messages.
+  ///
+  /// @param limit_batches the consumer will only consume the given number of batches.
+  /// @param timeout the maximum duration during which the command should run.
+  ///
+  /// @throws ConsumerRunningException if the consumer is already running
+  /// @throws ConsumerStartFailedException if the commited offsets cannot be restored
+  void StartWithLimit(uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const;
 
   /// Stops consuming messages.
   ///
@@ -136,9 +144,9 @@ class Consumer final : public RdKafka::EventCb {
   ///                      used.
   /// @param check_consumer_function a function to feed the received messages in, only used during this dry-run.
   ///
-  /// @throws ConsumerRunningException if the consumer is alredy running.
+  /// @throws ConsumerRunningException if the consumer is already running.
   /// @throws ConsumerCheckFailedException if check isn't successful.
-  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> limit_batches,
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> limit_batches,
              const ConsumerFunction &check_consumer_function) const;
 
   /// Returns true if the consumer is actively consuming messages.
@@ -157,6 +165,7 @@ class Consumer final : public RdKafka::EventCb {
   void event_cb(RdKafka::Event &event) override;
 
   void StartConsuming();
+  void StartConsumingWithLimit(uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const;
 
   void StopConsuming();
 
@@ -178,7 +187,6 @@ class Consumer final : public RdKafka::EventCb {
   ConsumerFunction consumer_function_;
   mutable std::atomic<bool> is_running_{false};
   mutable std::vector<RdKafka::TopicPartition *> last_assignment_;  // Protected by is_running_
-  std::optional<int64_t> limit_batches_{std::nullopt};
   std::unique_ptr<RdKafka::KafkaConsumer, std::function<void(RdKafka::KafkaConsumer *)>> consumer_;
   std::thread thread_;
   ConsumerRebalanceCb cb_;

src/integrations/kafka/exceptions.hpp

@@ -64,4 +64,16 @@ class TopicNotFoundException : public KafkaStreamException {
   TopicNotFoundException(const std::string_view consumer_name, const std::string_view topic_name)
       : KafkaStreamException("Kafka consumer {} cannot find topic {}", consumer_name, topic_name) {}
 };
+
+class ConsumerCommitFailedException : public KafkaStreamException {
+ public:
+  ConsumerCommitFailedException(const std::string_view consumer_name, const std::string_view error)
+      : KafkaStreamException("Committing offset of consumer {} failed: {}", consumer_name, error) {}
+};
+
+class ConsumerReadMessagesFailedException : public KafkaStreamException {
+ public:
+  ConsumerReadMessagesFailedException(const std::string_view consumer_name, const std::string_view error)
+      : KafkaStreamException("Error happened in consumer {} while fetching messages: {}", consumer_name, error) {}
+};
 }  // namespace memgraph::integrations::kafka

src/integrations/pulsar/consumer.cpp

@@ -11,13 +11,14 @@
 
 #include "integrations/pulsar/consumer.hpp"
 
+#include <algorithm>
+#include <chrono>
+#include <thread>
+
 #include <fmt/format.h>
 #include <pulsar/Client.h>
 #include <pulsar/InitialPosition.h>
 
-#include <chrono>
-#include <thread>
-
 #include "integrations/constants.hpp"
 #include "integrations/pulsar/exceptions.hpp"
 #include "utils/concepts.hpp"
@@ -33,6 +34,10 @@ namespace {
 template <typename T>
 concept PulsarConsumer = utils::SameAsAnyOf<T, pulsar_client::Consumer, pulsar_client::Reader>;
 
+template <typename TFunc>
+concept PulsarMessageGetter =
+    std::same_as<const pulsar_client::Message &, std::invoke_result_t<TFunc, const Message &>>;
+
 pulsar_client::Result ConsumeMessage(pulsar_client::Consumer &consumer, pulsar_client::Message &message,
                                      int remaining_timeout_in_ms) {
   return consumer.receive(message, remaining_timeout_in_ms);
@@ -97,6 +102,26 @@ pulsar_client::Client CreateClient(const std::string &service_url) {
   conf.setLogger(new SpdlogLoggerFactory);
   return {service_url, conf};
 }
+
+template <PulsarConsumer TConsumer, PulsarMessageGetter TPulsarMessageGetter>
+void TryToConsumeBatch(TConsumer &consumer, const ConsumerInfo &info, const ConsumerFunction &consumer_function,
+                       pulsar_client::MessageId &last_message_id, const std::vector<Message> &batch,
+                       const TPulsarMessageGetter &message_getter) {
+  consumer_function(batch);
+
+  auto has_message_failed = [&consumer, &info, &last_message_id, &message_getter](const auto &message) {
+    if (const auto result = consumer.acknowledge(message_getter(message)); result != pulsar_client::ResultOk) {
+      spdlog::warn("Acknowledging a message of consumer {} failed: {}", info.consumer_name, result);
+      return true;
+    }
+    last_message_id = message_getter(message).getMessageId();
+    return false;
+  };
+
+  if (std::ranges::any_of(batch, has_message_failed)) {
+    throw ConsumerAcknowledgeMessagesFailedException(info.consumer_name);
+  }
+}
 }  // namespace
 
 Message::Message(pulsar_client::Message &&message) : message_{std::move(message)} {}
@@ -137,6 +162,24 @@ void Consumer::Start() {
   StartConsuming();
 }
 
+void Consumer::StartWithLimit(const uint64_t limit_batches,
+                              const std::optional<std::chrono::milliseconds> timeout) const {
+  if (is_running_) {
+    throw ConsumerRunningException(info_.consumer_name);
+  }
+  if (limit_batches < kMinimumStartBatchLimit) {
+    throw ConsumerStartFailedException(
+        info_.consumer_name, fmt::format("Batch limit has to be greater than or equal to {}", kMinimumStartBatchLimit));
+  }
+  if (timeout.value_or(kMinimumInterval) < kMinimumInterval) {
+    throw ConsumerStartFailedException(
+        info_.consumer_name,
+        fmt::format("Timeout has to be greater than or equal to {} milliseconds", kMinimumInterval.count()));
+  }
+
+  StartConsumingWithLimit(limit_batches, timeout);
+}
+
 void Consumer::Stop() {
   if (!is_running_) {
     throw ConsumerStoppedException(info_.consumer_name);
@@ -154,7 +197,7 @@ void Consumer::StopIfRunning() {
   }
 }
 
-void Consumer::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> limit_batches,
+void Consumer::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> limit_batches,
                      const ConsumerFunction &check_consumer_function) const {
   // NOLINTNEXTLINE (modernize-use-nullptr)
   if (timeout.value_or(kMinimumInterval) < kMinimumInterval) {
@@ -240,9 +283,7 @@ void Consumer::StartConsuming() {
       auto maybe_batch = GetBatch(consumer_, info_, is_running_, last_message_id_);
 
       if (maybe_batch.HasError()) {
-        spdlog::warn("Error happened in consumer {} while fetching messages: {}!", info_.consumer_name,
-                     maybe_batch.GetError());
-        break;
+        throw ConsumerReadMessagesFailedException(info_.consumer_name, maybe_batch.GetError());
       }
 
       const auto &batch = maybe_batch.GetValue();
@@ -254,18 +295,8 @@ void Consumer::StartConsuming() {
       spdlog::info("Pulsar consumer {} is processing a batch", info_.consumer_name);
 
       try {
-        consumer_function_(batch);
-
-        if (std::any_of(batch.begin(), batch.end(), [&](const auto &message) {
-              if (const auto result = consumer_.acknowledge(message.message_); result != pulsar_client::ResultOk) {
-                spdlog::warn("Acknowledging a message of consumer {} failed: {}", info_.consumer_name, result);
-                return true;
-              }
-              last_message_id_ = message.message_.getMessageId();
-              return false;
-            })) {
-          break;
-        }
+        TryToConsumeBatch(consumer_, info_, consumer_function_, last_message_id_, batch,
+                          [&](const Message &message) -> const pulsar_client::Message & { return message.message_; });
       } catch (const std::exception &e) {
         spdlog::warn("Error happened in consumer {} while processing a batch: {}!", info_.consumer_name, e.what());
         break;
@@ -277,6 +308,43 @@ void Consumer::StartConsuming() {
   });
 }
 
+void Consumer::StartConsumingWithLimit(uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const {
+  if (is_running_.exchange(true)) {
+    throw ConsumerRunningException(info_.consumer_name);
+  }
+  utils::OnScopeExit restore_is_running([this] { is_running_.store(false); });
+
+  const auto timeout_to_use = timeout.value_or(kDefaultCheckTimeout);
+  const auto start = std::chrono::steady_clock::now();
+
+  for (uint64_t batch_count = 0; batch_count < limit_batches;) {
+    const auto now = std::chrono::steady_clock::now();
+    if (now - start >= timeout_to_use) {
+      throw ConsumerCheckFailedException(info_.consumer_name, "Timeout reached");
+    }
+
+    const auto maybe_batch = GetBatch(consumer_, info_, is_running_, last_message_id_);
+
+    if (maybe_batch.HasError()) {
+      throw ConsumerReadMessagesFailedException(info_.consumer_name, maybe_batch.GetError());
+    }
+
+    const auto &batch = maybe_batch.GetValue();
+
+    if (batch.empty()) {
+      continue;
+    }
+    ++batch_count;
+
+    spdlog::info("Pulsar consumer {} is processing a batch", info_.consumer_name);
+
+    TryToConsumeBatch(consumer_, info_, consumer_function_, last_message_id_, batch,
+                      [](const Message &message) -> const pulsar_client::Message & { return message.message_; });
+
+    spdlog::info("Pulsar consumer {} finished processing", info_.consumer_name);
+  }
+}
+
 void Consumer::StopConsuming() {
   is_running_.store(false);
   if (thread_.joinable()) {

src/integrations/pulsar/consumer.hpp

@@ -58,25 +58,27 @@ class Consumer final {
 
   bool IsRunning() const;
   void Start();
+  void StartWithLimit(uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const;
   void Stop();
   void StopIfRunning();
 
-  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> limit_batches,
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> limit_batches,
              const ConsumerFunction &check_consumer_function) const;
 
   const ConsumerInfo &Info() const;
 
  private:
   void StartConsuming();
+  void StartConsumingWithLimit(uint64_t limit_batches, std::optional<std::chrono::milliseconds> timeout) const;
   void StopConsuming();
 
   ConsumerInfo info_;
   mutable pulsar_client::Client client_;
-  pulsar_client::Consumer consumer_;
+  mutable pulsar_client::Consumer consumer_;
   ConsumerFunction consumer_function_;
 
   mutable std::atomic<bool> is_running_{false};
-  pulsar_client::MessageId last_message_id_{pulsar_client::MessageId::earliest()};
+  mutable pulsar_client::MessageId last_message_id_{pulsar_client::MessageId::earliest()};  // Protected by is_running_
   std::thread thread_;
 };
 }  // namespace memgraph::integrations::pulsar

src/integrations/pulsar/exceptions.hpp

@@ -55,4 +55,16 @@ class TopicNotFoundException : public PulsarStreamException {
   TopicNotFoundException(const std::string &consumer_name, const std::string &topic_name)
       : PulsarStreamException("Pulsar consumer {} cannot find topic {}", consumer_name, topic_name) {}
 };
+
+class ConsumerReadMessagesFailedException : public PulsarStreamException {
+ public:
+  ConsumerReadMessagesFailedException(const std::string_view consumer_name, const std::string_view error)
+      : PulsarStreamException("Error happened in consumer {} while fetching messages: {}", consumer_name, error) {}
+};
+
+class ConsumerAcknowledgeMessagesFailedException : public PulsarStreamException {
+ public:
+  explicit ConsumerAcknowledgeMessagesFailedException(const std::string_view consumer_name)
+      : PulsarStreamException("Acknowledging a message of consumer {} has failed!", consumer_name) {}
+};
 }  // namespace memgraph::integrations::pulsar

src/memgraph.cpp

@@ -216,6 +216,11 @@ DEFINE_bool(telemetry_enabled, false,
             "the database runtime (vertex and edge counts and resource usage) "
             "to allow for easier improvement of the product.");
 
+// NOLINTNEXTLINE(cppcoreguidelines-avoid-non-const-global-variables)
+DEFINE_bool(storage_restore_replicas_on_startup, true,
+            "Controls replicas should be restored automatically.");  // TODO(42jeremy) this must be removed once T0835
+                                                                     // is implemented.
+
 // Streams flags
 // NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
 DEFINE_uint32(
@@ -252,6 +257,11 @@ DEFINE_double(query_execution_timeout_sec, 600,
               "Maximum allowed query execution time. Queries exceeding this "
               "limit will be aborted. Value of 0 means no limit.");
 
+// NOLINTNEXTLINE(cppcoreguidelines-avoid-non-const-global-variables)
+DEFINE_uint64(replication_replica_check_frequency_sec, 1,
+              "The time duration between two replica checks/pings. If < 1, replicas will NOT be checked at all. NOTE: "
+              "The MAIN instance allocates a new thread for each REPLICA.");
+
 // NOLINTNEXTLINE (cppcoreguidelines-avoid-non-const-global-variables)
 DEFINE_uint64(
     memory_limit, 0,
@@ -1070,6 +1080,22 @@ int main(int argc, char **argv) {
       if (maybe_exc) {
         spdlog::error(memgraph::utils::MessageWithLink("Unable to load support for embedded Python: {}.", *maybe_exc,
                                                        "https://memgr.ph/python"));
+      } else {
+        // Change how we load dynamic libraries on Python by using RTLD_NOW and
+        // RTLD_DEEPBIND flags. This solves an issue with using the wrong version of
+        // libstd.
+        auto gil = memgraph::py::EnsureGIL();
+        // NOLINTNEXTLINE(hicpp-signed-bitwise)
+        auto *flag = PyLong_FromLong(RTLD_NOW | RTLD_DEEPBIND);
+        auto *setdl = PySys_GetObject("setdlopenflags");
+        MG_ASSERT(setdl);
+        auto *arg = PyTuple_New(1);
+        MG_ASSERT(arg);
+        MG_ASSERT(PyTuple_SetItem(arg, 0, flag) == 0);
+        PyObject_CallObject(setdl, arg);
+        Py_DECREF(flag);
+        Py_DECREF(setdl);
+        Py_DECREF(arg);
       }
     } else {
       spdlog::error(
@@ -1174,7 +1200,8 @@ int main(int argc, char **argv) {
                      .snapshot_retention_count = FLAGS_storage_snapshot_retention_count,
                      .wal_file_size_kibibytes = FLAGS_storage_wal_file_size_kib,
                      .wal_file_flush_every_n_tx = FLAGS_storage_wal_file_flush_every_n_tx,
-                     .snapshot_on_exit = FLAGS_storage_snapshot_on_exit},
+                     .snapshot_on_exit = FLAGS_storage_snapshot_on_exit,
+                     .restore_replicas_on_startup = FLAGS_storage_restore_replicas_on_startup},
       .transaction = {.isolation_level = ParseIsolationLevel()}};
   if (FLAGS_storage_snapshot_interval_sec == 0) {
     if (FLAGS_storage_wal_enabled) {
@@ -1200,6 +1227,7 @@ int main(int argc, char **argv) {
       &db,
       {.query = {.allow_load_csv = FLAGS_allow_load_csv},
        .execution_timeout_sec = FLAGS_query_execution_timeout_sec,
+       .replication_replica_check_frequency = std::chrono::seconds(FLAGS_replication_replica_check_frequency_sec),
        .default_kafka_bootstrap_servers = FLAGS_kafka_bootstrap_servers,
        .default_pulsar_service_url = FLAGS_pulsar_service_url,
        .stream_transaction_conflict_retries = FLAGS_stream_transaction_conflict_retries,

src/memory/CMakeLists.txt

@@ -9,4 +9,5 @@ target_link_libraries(mg-memory mg-utils fmt)
 
 if (ENABLE_JEMALLOC)
   target_link_libraries(mg-memory Jemalloc::Jemalloc)
+  target_compile_definitions(mg-memory PRIVATE USE_JEMALLOC=1)
 endif()

src/mg_import_csv.cpp

@@ -170,7 +170,7 @@ enum class CsvParserState {
   EXPECT_DELIMITER,
 };
 
-bool SubstringStartsWith(const std::string_view &str, size_t pos, const std::string_view &what) {
+bool SubstringStartsWith(const std::string_view str, size_t pos, const std::string_view what) {
   return memgraph::utils::StartsWith(memgraph::utils::Substr(str, pos), what);
 }
 

src/query/config.hpp

@@ -21,6 +21,8 @@ struct InterpreterConfig {
 
   // The default execution timeout is 10 minutes.
   double execution_timeout_sec{600.0};
+  // The same as \ref memgraph::storage::replication::ReplicationClientConfig
+  std::chrono::seconds replication_replica_check_frequency{1};
 
   std::string default_kafka_bootstrap_servers;
   std::string default_pulsar_service_url;

src/query/db_accessor.hpp

@@ -310,11 +310,11 @@ class DbAccessor final {
     return std::make_optional<VertexAccessor>(*value);
   }
 
-  storage::PropertyId NameToProperty(const std::string_view &name) { return accessor_->NameToProperty(name); }
+  storage::PropertyId NameToProperty(const std::string_view name) { return accessor_->NameToProperty(name); }
 
-  storage::LabelId NameToLabel(const std::string_view &name) { return accessor_->NameToLabel(name); }
+  storage::LabelId NameToLabel(const std::string_view name) { return accessor_->NameToLabel(name); }
 
-  storage::EdgeTypeId NameToEdgeType(const std::string_view &name) { return accessor_->NameToEdgeType(name); }
+  storage::EdgeTypeId NameToEdgeType(const std::string_view name) { return accessor_->NameToEdgeType(name); }
 
   const std::string &PropertyToName(storage::PropertyId prop) const { return accessor_->PropertyToName(prop); }
 

src/query/dump.cpp

@@ -46,7 +46,7 @@ const char *kInternalPropertyId = "__mg_id__";
 const char *kInternalVertexLabel = "__mg_vertex__";
 
 /// A helper function that escapes label, edge type and property names.
-std::string EscapeName(const std::string_view &value) {
+std::string EscapeName(const std::string_view value) {
   std::string out;
   out.reserve(value.size() + 2);
   out.append(1, '`');

src/query/frontend/ast/ast.lcp

@@ -2375,10 +2375,7 @@ cpp<#
    (port "Expression *" :initval "nullptr" :scope :public
              :slk-save #'slk-save-ast-pointer
              :slk-load (slk-load-ast-pointer "Expression"))
-   (sync_mode "SyncMode" :scope :public)
-   (timeout "Expression *" :initval "nullptr" :scope :public
-             :slk-save #'slk-save-ast-pointer
-             :slk-load (slk-load-ast-pointer "Expression")))
+   (sync_mode "SyncMode" :scope :public))
 
   (:public
     (lcp:define-enum action
@@ -2391,6 +2388,9 @@ cpp<#
     (lcp:define-enum sync-mode
         (sync async)
       (:serialize))
+    (lcp:define-enum replica-state
+        (ready replicating recovery invalid)
+      (:serialize))
     #>cpp
     ReplicationQuery() = default;
 

src/query/frontend/ast/cypher_main_visitor.cpp

@@ -275,18 +275,7 @@ antlrcpp::Any CypherMainVisitor::visitRegisterReplica(MemgraphCypher::RegisterRe
   replication_query->replica_name_ = ctx->replicaName()->symbolicName()->accept(this).as<std::string>();
   if (ctx->SYNC()) {
     replication_query->sync_mode_ = memgraph::query::ReplicationQuery::SyncMode::SYNC;
-    if (ctx->WITH() && ctx->TIMEOUT()) {
-      if (ctx->timeout->numberLiteral()) {
-        // we accept both double and integer literals
-        replication_query->timeout_ = ctx->timeout->accept(this);
-      } else {
-        throw SemanticException("Timeout should be a integer or double literal!");
-      }
-    }
   } else if (ctx->ASYNC()) {
-    if (ctx->WITH() && ctx->TIMEOUT()) {
-      throw SyntaxException("Timeout can be set only for the SYNC replication mode!");
-    }
     replication_query->sync_mode_ = memgraph::query::ReplicationQuery::SyncMode::ASYNC;
   }
 
@@ -775,6 +764,23 @@ antlrcpp::Any CypherMainVisitor::visitDropStream(MemgraphCypher::DropStreamConte
 antlrcpp::Any CypherMainVisitor::visitStartStream(MemgraphCypher::StartStreamContext *ctx) {
   auto *stream_query = storage_->Create<StreamQuery>();
   stream_query->action_ = StreamQuery::Action::START_STREAM;
+
+  if (ctx->BATCH_LIMIT()) {
+    if (!ctx->batchLimit->numberLiteral() || !ctx->batchLimit->numberLiteral()->integerLiteral()) {
+      throw SemanticException("Batch limit should be an integer literal!");
+    }
+    stream_query->batch_limit_ = ctx->batchLimit->accept(this);
+  }
+  if (ctx->TIMEOUT()) {
+    if (!ctx->timeout->numberLiteral() || !ctx->timeout->numberLiteral()->integerLiteral()) {
+      throw SemanticException("Timeout should be an integer literal!");
+    }
+    if (!ctx->BATCH_LIMIT()) {
+      throw SemanticException("Parameter TIMEOUT can only be defined if BATCH_LIMIT is defined");
+    }
+    stream_query->timeout_ = ctx->timeout->accept(this);
+  }
+
   stream_query->stream_name_ = ctx->streamName()->symbolicName()->accept(this).as<std::string>();
   return stream_query;
 }

src/query/frontend/opencypher/grammar/MemgraphCypher.g4

@@ -276,7 +276,6 @@ replicaName : symbolicName ;
 socketAddress : literal ;
 
 registerReplica : REGISTER REPLICA replicaName ( SYNC | ASYNC )
-                ( WITH TIMEOUT timeout=literal ) ?
                 TO socketAddress ;
 
 dropReplica : DROP REPLICA replicaName ;
@@ -351,7 +350,7 @@ pulsarCreateStream : CREATE PULSAR STREAM streamName ( pulsarCreateStreamConfig
 
 dropStream : DROP STREAM streamName ;
 
-startStream : START STREAM streamName ;
+startStream : START STREAM streamName ( BATCH_LIMIT batchLimit=literal ) ? ( TIMEOUT timeout=literal ) ? ;
 
 startAllStreams : START ALL STREAMS ;
 

src/query/interpret/awesome_memgraph_functions.cpp

@@ -882,21 +882,36 @@ TypedValue Id(const TypedValue *args, int64_t nargs, const FunctionContext &ctx)
 }
 
 TypedValue ToString(const TypedValue *args, int64_t nargs, const FunctionContext &ctx) {
-  FType<Or<Null, String, Number, Bool>>("toString", args, nargs);
+  FType<Or<Null, String, Number, Date, LocalTime, LocalDateTime, Duration, Bool>>("toString", args, nargs);
   const auto &arg = args[0];
   if (arg.IsNull()) {
     return TypedValue(ctx.memory);
-  } else if (arg.IsString()) {
+  }
+  if (arg.IsString()) {
     return TypedValue(arg, ctx.memory);
-  } else if (arg.IsInt()) {
+  }
+  if (arg.IsInt()) {
     // TODO: This is making a pointless copy of std::string, we may want to
     // use a different conversion to string
     return TypedValue(std::to_string(arg.ValueInt()), ctx.memory);
-  } else if (arg.IsDouble()) {
-    return TypedValue(std::to_string(arg.ValueDouble()), ctx.memory);
-  } else {
-    return TypedValue(arg.ValueBool() ? "true" : "false", ctx.memory);
   }
+  if (arg.IsDouble()) {
+    return TypedValue(std::to_string(arg.ValueDouble()), ctx.memory);
+  }
+  if (arg.IsDate()) {
+    return TypedValue(arg.ValueDate().ToString(), ctx.memory);
+  }
+  if (arg.IsLocalTime()) {
+    return TypedValue(arg.ValueLocalTime().ToString(), ctx.memory);
+  }
+  if (arg.IsLocalDateTime()) {
+    return TypedValue(arg.ValueLocalDateTime().ToString(), ctx.memory);
+  }
+  if (arg.IsDuration()) {
+    return TypedValue(arg.ValueDuration().ToString(), ctx.memory);
+  }
+
+  return TypedValue(arg.ValueBool() ? "true" : "false", ctx.memory);
 }
 
 TypedValue Timestamp(const TypedValue *args, int64_t nargs, const FunctionContext &ctx) {

src/query/interpret/eval.hpp

@@ -111,7 +111,7 @@ class ExpressionEvaluator : public ExpressionVisitor<TypedValue> {
   TypedValue Visit(IfOperator &if_operator) override {
     auto condition = if_operator.condition_->Accept(*this);
     if (condition.IsNull()) {
-      return if_operator.then_expression_->Accept(*this);
+      return if_operator.else_expression_->Accept(*this);
     }
     if (condition.type() != TypedValue::Type::Bool) {
       // At the moment IfOperator is used only in CASE construct.
@@ -716,7 +716,7 @@ class ExpressionEvaluator : public ExpressionVisitor<TypedValue> {
   }
 
   template <class TRecordAccessor>
-  storage::PropertyValue GetProperty(const TRecordAccessor &record_accessor, const std::string_view &name) {
+  storage::PropertyValue GetProperty(const TRecordAccessor &record_accessor, const std::string_view name) {
     auto maybe_prop = record_accessor.GetProperty(view_, dba_->NameToProperty(name));
     if (maybe_prop.HasError() && maybe_prop.GetError() == storage::Error::NONEXISTENT_OBJECT) {
       // This is a very nasty and temporary hack in order to make MERGE work.

src/query/interpreter.cpp

@@ -44,6 +44,7 @@
 #include "query/trigger.hpp"
 #include "query/typed_value.hpp"
 #include "storage/v2/property_value.hpp"
+#include "storage/v2/replication/enums.hpp"
 #include "utils/algorithm.hpp"
 #include "utils/csv_parsing.hpp"
 #include "utils/event_counter.hpp"
@@ -160,7 +161,8 @@ class ReplQueryHandler final : public query::ReplicationQueryHandler {
 
   /// @throw QueryRuntimeException if an error ocurred.
   void RegisterReplica(const std::string &name, const std::string &socket_address,
-                       const ReplicationQuery::SyncMode sync_mode, const std::optional<double> timeout) override {
+                       const ReplicationQuery::SyncMode sync_mode,
+                       const std::chrono::seconds replica_check_frequency) override {
     if (db_->GetReplicationRole() == storage::ReplicationRole::REPLICA) {
       // replica can't register another replica
       throw QueryRuntimeException("Replica can't register another replica!");
@@ -182,8 +184,9 @@ class ReplQueryHandler final : public query::ReplicationQueryHandler {
         io::network::Endpoint::ParseSocketOrIpAddress(socket_address, query::kDefaultReplicationPort);
     if (maybe_ip_and_port) {
       auto [ip, port] = *maybe_ip_and_port;
-      auto ret =
-          db_->RegisterReplica(name, {std::move(ip), port}, repl_mode, {.timeout = timeout, .ssl = std::nullopt});
+      auto ret = db_->RegisterReplica(name, {std::move(ip), port}, repl_mode,
+                                      storage::replication::RegistrationMode::MUST_BE_INSTANTLY_VALID,
+                                      {.replica_check_frequency = replica_check_frequency, .ssl = std::nullopt});
       if (ret.HasError()) {
         throw QueryRuntimeException(fmt::format("Couldn't register replica '{}'!", name));
       }
@@ -226,8 +229,24 @@ class ReplQueryHandler final : public query::ReplicationQueryHandler {
           replica.sync_mode = ReplicationQuery::SyncMode::ASYNC;
           break;
       }
-      if (repl_info.timeout) {
-        replica.timeout = *repl_info.timeout;
+
+      replica.current_timestamp_of_replica = repl_info.timestamp_info.current_timestamp_of_replica;
+      replica.current_number_of_timestamp_behind_master =
+          repl_info.timestamp_info.current_number_of_timestamp_behind_master;
+
+      switch (repl_info.state) {
+        case storage::replication::ReplicaState::READY:
+          replica.state = ReplicationQuery::ReplicaState::READY;
+          break;
+        case storage::replication::ReplicaState::REPLICATING:
+          replica.state = ReplicationQuery::ReplicaState::REPLICATING;
+          break;
+        case storage::replication::ReplicaState::RECOVERY:
+          replica.state = ReplicationQuery::ReplicaState::RECOVERY;
+          break;
+        case storage::replication::ReplicaState::INVALID:
+          replica.state = ReplicationQuery::ReplicaState::INVALID;
+          break;
       }
 
       return replica;
@@ -448,7 +467,7 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
       return callback;
     }
     case ReplicationQuery::Action::SHOW_REPLICATION_ROLE: {
-      callback.header = {"replication mode"};
+      callback.header = {"replication role"};
       callback.fn = [handler = ReplQueryHandler{interpreter_context->db}] {
         auto mode = handler.ShowReplicationRole();
         switch (mode) {
@@ -466,22 +485,18 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
       const auto &name = repl_query->replica_name_;
       const auto &sync_mode = repl_query->sync_mode_;
       auto socket_address = repl_query->socket_address_->Accept(evaluator);
-      auto timeout = EvaluateOptionalExpression(repl_query->timeout_, &evaluator);
-      std::optional<double> maybe_timeout;
-      if (timeout.IsDouble()) {
-        maybe_timeout = timeout.ValueDouble();
-      } else if (timeout.IsInt()) {
-        maybe_timeout = static_cast<double>(timeout.ValueInt());
-      }
+      const auto replica_check_frequency = interpreter_context->config.replication_replica_check_frequency;
+
       callback.fn = [handler = ReplQueryHandler{interpreter_context->db}, name, socket_address, sync_mode,
-                     maybe_timeout]() mutable {
-        handler.RegisterReplica(name, std::string(socket_address.ValueString()), sync_mode, maybe_timeout);
+                     replica_check_frequency]() mutable {
+        handler.RegisterReplica(name, std::string(socket_address.ValueString()), sync_mode, replica_check_frequency);
         return std::vector<std::vector<TypedValue>>();
       };
       notifications->emplace_back(SeverityLevel::INFO, NotificationCode::REGISTER_REPLICA,
                                   fmt::format("Replica {} is registered.", repl_query->replica_name_));
       return callback;
     }
+
     case ReplicationQuery::Action::DROP_REPLICA: {
       const auto &name = repl_query->replica_name_;
       callback.fn = [handler = ReplQueryHandler{interpreter_context->db}, name]() mutable {
@@ -492,8 +507,11 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
                                   fmt::format("Replica {} is dropped.", repl_query->replica_name_));
       return callback;
     }
+
     case ReplicationQuery::Action::SHOW_REPLICAS: {
-      callback.header = {"name", "socket_address", "sync_mode", "timeout"};
+      callback.header = {
+          "name", "socket_address", "sync_mode", "current_timestamp_of_replica", "number_of_timestamp_behind_master",
+          "state"};
       callback.fn = [handler = ReplQueryHandler{interpreter_context->db}, replica_nfields = callback.header.size()] {
         const auto &replicas = handler.ShowReplicas();
         auto typed_replicas = std::vector<std::vector<TypedValue>>{};
@@ -504,6 +522,7 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
 
           typed_replica.emplace_back(TypedValue(replica.name));
           typed_replica.emplace_back(TypedValue(replica.socket_address));
+
           switch (replica.sync_mode) {
             case ReplicationQuery::SyncMode::SYNC:
               typed_replica.emplace_back(TypedValue("sync"));
@@ -512,11 +531,24 @@ Callback HandleReplicationQuery(ReplicationQuery *repl_query, const Parameters &
               typed_replica.emplace_back(TypedValue("async"));
               break;
           }
-          typed_replica.emplace_back(TypedValue(static_cast<int64_t>(replica.sync_mode)));
-          if (replica.timeout) {
-            typed_replica.emplace_back(TypedValue(*replica.timeout));
-          } else {
-            typed_replica.emplace_back(TypedValue());
+
+          typed_replica.emplace_back(TypedValue(static_cast<int64_t>(replica.current_timestamp_of_replica)));
+          typed_replica.emplace_back(
+              TypedValue(static_cast<int64_t>(replica.current_number_of_timestamp_behind_master)));
+
+          switch (replica.state) {
+            case ReplicationQuery::ReplicaState::READY:
+              typed_replica.emplace_back(TypedValue("ready"));
+              break;
+            case ReplicationQuery::ReplicaState::REPLICATING:
+              typed_replica.emplace_back(TypedValue("replicating"));
+              break;
+            case ReplicationQuery::ReplicaState::RECOVERY:
+              typed_replica.emplace_back(TypedValue("recovery"));
+              break;
+            case ReplicationQuery::ReplicaState::INVALID:
+              typed_replica.emplace_back(TypedValue("invalid"));
+              break;
           }
 
           typed_replicas.emplace_back(std::move(typed_replica));
@@ -652,12 +684,26 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
       return callback;
     }
     case StreamQuery::Action::START_STREAM: {
-      callback.fn = [interpreter_context, stream_name = stream_query->stream_name_]() {
-        interpreter_context->streams.Start(stream_name);
-        return std::vector<std::vector<TypedValue>>{};
-      };
-      notifications->emplace_back(SeverityLevel::INFO, NotificationCode::START_STREAM,
-                                  fmt::format("Started stream {}.", stream_query->stream_name_));
+      const auto batch_limit = GetOptionalValue<int64_t>(stream_query->batch_limit_, evaluator);
+      const auto timeout = GetOptionalValue<std::chrono::milliseconds>(stream_query->timeout_, evaluator);
+
+      if (batch_limit.has_value()) {
+        if (batch_limit.value() < 0) {
+          throw utils::BasicException("Parameter BATCH_LIMIT cannot hold negative value");
+        }
+
+        callback.fn = [interpreter_context, stream_name = stream_query->stream_name_, batch_limit, timeout]() {
+          interpreter_context->streams.StartWithLimit(stream_name, static_cast<uint64_t>(batch_limit.value()), timeout);
+          return std::vector<std::vector<TypedValue>>{};
+        };
+      } else {
+        callback.fn = [interpreter_context, stream_name = stream_query->stream_name_]() {
+          interpreter_context->streams.Start(stream_name);
+          return std::vector<std::vector<TypedValue>>{};
+        };
+        notifications->emplace_back(SeverityLevel::INFO, NotificationCode::START_STREAM,
+                                    fmt::format("Started stream {}.", stream_query->stream_name_));
+      }
       return callback;
     }
     case StreamQuery::Action::START_ALL_STREAMS: {
@@ -726,10 +772,16 @@ Callback HandleStreamQuery(StreamQuery *stream_query, const Parameters &paramete
       return callback;
     }
     case StreamQuery::Action::CHECK_STREAM: {
-      callback.header = {"query", "parameters"};
+      callback.header = {"queries", "raw messages"};
+
+      const auto batch_limit = GetOptionalValue<int64_t>(stream_query->batch_limit_, evaluator);
+      if (batch_limit.has_value() && batch_limit.value() < 0) {
+        throw utils::BasicException("Parameter BATCH_LIMIT cannot hold negative value");
+      }
+
       callback.fn = [interpreter_context, stream_name = stream_query->stream_name_,
                      timeout = GetOptionalValue<std::chrono::milliseconds>(stream_query->timeout_, evaluator),
-                     batch_limit = GetOptionalValue<int64_t>(stream_query->batch_limit_, evaluator)]() mutable {
+                     batch_limit]() mutable {
         return interpreter_context->streams.Check(stream_name, timeout, batch_limit);
       };
       notifications->emplace_back(SeverityLevel::INFO, NotificationCode::CHECK_STREAM,

src/query/interpreter.hpp

@@ -127,6 +127,9 @@ class ReplicationQueryHandler {
     std::string socket_address;
     ReplicationQuery::SyncMode sync_mode;
     std::optional<double> timeout;
+    uint64_t current_timestamp_of_replica;
+    uint64_t current_number_of_timestamp_behind_master;
+    ReplicationQuery::ReplicaState state;
   };
 
   /// @throw QueryRuntimeException if an error ocurred.
@@ -137,7 +140,8 @@ class ReplicationQueryHandler {
 
   /// @throw QueryRuntimeException if an error ocurred.
   virtual void RegisterReplica(const std::string &name, const std::string &socket_address,
-                               const ReplicationQuery::SyncMode sync_mode, const std::optional<double> timeout) = 0;
+                               ReplicationQuery::SyncMode sync_mode,
+                               const std::chrono::seconds replica_check_frequency) = 0;
 
   /// @throw QueryRuntimeException if an error ocurred.
   virtual void DropReplica(const std::string &replica_name) = 0;

src/query/plan/operator.cpp

@@ -3697,7 +3697,7 @@ std::unordered_map<std::string, int64_t> CallProcedure::GetAndResetCounters() {
 
 namespace {
 
-void CallCustomProcedure(const std::string_view &fully_qualified_procedure_name, const mgp_proc &proc,
+void CallCustomProcedure(const std::string_view fully_qualified_procedure_name, const mgp_proc &proc,
                          const std::vector<Expression *> &args, mgp_graph &graph, ExpressionEvaluator *evaluator,
                          utils::MemoryResource *memory, std::optional<size_t> memory_limit, mgp_result *result) {
   static_assert(std::uses_allocator_v<mgp_value, utils::Allocator<mgp_value>>,

src/query/procedure/mg_procedure_helpers.cpp

@@ -24,7 +24,7 @@ MgpUniquePtr<mgp_value> GetStringValueOrSetError(const char *string, mgp_memory
 }
 
 bool InsertResultOrSetError(mgp_result *result, mgp_result_record *record, const char *result_name, mgp_value *value) {
-  if (const auto err = mgp_result_record_insert(record, result_name, value); err != MGP_ERROR_NO_ERROR) {
+  if (const auto err = mgp_result_record_insert(record, result_name, value); err != mgp_error::MGP_ERROR_NO_ERROR) {
     const auto error_msg = fmt::format("Unable to set the result for {}, error = {}", result_name, err);
     static_cast<void>(mgp_result_set_error_msg(result, error_msg.c_str()));
     return false;

src/query/procedure/mg_procedure_helpers.hpp

@@ -25,7 +25,7 @@ TResult Call(TFunc func, TArgs... args) {
   static_assert(std::is_trivially_copyable_v<TFunc>);
   static_assert((std::is_trivially_copyable_v<std::remove_reference_t<TArgs>> && ...));
   TResult result{};
-  MG_ASSERT(func(args..., &result) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(func(args..., &result) == mgp_error::MGP_ERROR_NO_ERROR);
   return result;
 }
 
@@ -50,10 +50,10 @@ mgp_error CreateMgpObject(MgpUniquePtr<TObj> &obj, TFunc func, TArgs &&...args)
 
 template <typename Fun>
 [[nodiscard]] bool TryOrSetError(Fun &&func, mgp_result *result) {
-  if (const auto err = func(); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+  if (const auto err = func(); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
     static_cast<void>(mgp_result_set_error_msg(result, "Not enough memory!"));
     return false;
-  } else if (err != MGP_ERROR_NO_ERROR) {
+  } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
     const auto error_msg = fmt::format("Unexpected error ({})!", err);
     static_cast<void>(mgp_result_set_error_msg(result, error_msg.c_str()));
     return false;

src/query/procedure/mg_procedure_impl.cpp

@@ -143,48 +143,48 @@ template <typename TFunc, typename... Args>
     WrapExceptionsHelper(std::forward<TFunc>(func), std::forward<Args>(args)...);
   } catch (const DeletedObjectException &neoe) {
     spdlog::error("Deleted object error during mg API call: {}", neoe.what());
-    return MGP_ERROR_DELETED_OBJECT;
+    return mgp_error::MGP_ERROR_DELETED_OBJECT;
   } catch (const KeyAlreadyExistsException &kaee) {
     spdlog::error("Key already exists error during mg API call: {}", kaee.what());
-    return MGP_ERROR_KEY_ALREADY_EXISTS;
+    return mgp_error::MGP_ERROR_KEY_ALREADY_EXISTS;
   } catch (const InsufficientBufferException &ibe) {
     spdlog::error("Insufficient buffer error during mg API call: {}", ibe.what());
-    return MGP_ERROR_INSUFFICIENT_BUFFER;
+    return mgp_error::MGP_ERROR_INSUFFICIENT_BUFFER;
   } catch (const ImmutableObjectException &ioe) {
     spdlog::error("Immutable object error during mg API call: {}", ioe.what());
-    return MGP_ERROR_IMMUTABLE_OBJECT;
+    return mgp_error::MGP_ERROR_IMMUTABLE_OBJECT;
   } catch (const ValueConversionException &vce) {
     spdlog::error("Value converion error during mg API call: {}", vce.what());
-    return MGP_ERROR_VALUE_CONVERSION;
+    return mgp_error::MGP_ERROR_VALUE_CONVERSION;
   } catch (const SerializationException &se) {
     spdlog::error("Serialization error during mg API call: {}", se.what());
-    return MGP_ERROR_SERIALIZATION_ERROR;
+    return mgp_error::MGP_ERROR_SERIALIZATION_ERROR;
   } catch (const std::bad_alloc &bae) {
     spdlog::error("Memory allocation error during mg API call: {}", bae.what());
-    return MGP_ERROR_UNABLE_TO_ALLOCATE;
+    return mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE;
   } catch (const memgraph::utils::OutOfMemoryException &oome) {
     spdlog::error("Memory limit exceeded during mg API call: {}", oome.what());
-    return MGP_ERROR_UNABLE_TO_ALLOCATE;
+    return mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE;
   } catch (const std::out_of_range &oore) {
     spdlog::error("Out of range error during mg API call: {}", oore.what());
-    return MGP_ERROR_OUT_OF_RANGE;
+    return mgp_error::MGP_ERROR_OUT_OF_RANGE;
   } catch (const std::invalid_argument &iae) {
     spdlog::error("Invalid argument error during mg API call: {}", iae.what());
-    return MGP_ERROR_INVALID_ARGUMENT;
+    return mgp_error::MGP_ERROR_INVALID_ARGUMENT;
   } catch (const std::logic_error &lee) {
     spdlog::error("Logic error during mg API call: {}", lee.what());
-    return MGP_ERROR_LOGIC_ERROR;
+    return mgp_error::MGP_ERROR_LOGIC_ERROR;
   } catch (const std::exception &e) {
     spdlog::error("Unexpected error during mg API call: {}", e.what());
-    return MGP_ERROR_UNKNOWN_ERROR;
+    return mgp_error::MGP_ERROR_UNKNOWN_ERROR;
   } catch (const memgraph::utils::temporal::InvalidArgumentException &e) {
     spdlog::error("Invalid argument was sent to an mg API call for temporal types: {}", e.what());
-    return MGP_ERROR_INVALID_ARGUMENT;
+    return mgp_error::MGP_ERROR_INVALID_ARGUMENT;
   } catch (...) {
     spdlog::error("Unexpected error during mg API call");
-    return MGP_ERROR_UNKNOWN_ERROR;
+    return mgp_error::MGP_ERROR_UNKNOWN_ERROR;
   }
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 // Graph mutations
@@ -846,7 +846,7 @@ mgp_value_type MgpValueGetType(const mgp_value &val) noexcept { return val.type;
 mgp_error mgp_value_get_type(mgp_value *val, mgp_value_type *result) {
   static_assert(noexcept(MgpValueGetType(*val)));
   *result = MgpValueGetType(*val);
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 // NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
@@ -854,7 +854,7 @@ mgp_error mgp_value_get_type(mgp_value *val, mgp_value_type *result) {
   mgp_error mgp_value_is_##type_lowercase(mgp_value *val, int *result) { \
     static_assert(noexcept(MgpValueGetType(*val)));                      \
     *result = MgpValueGetType(*val) == MGP_VALUE_TYPE_##type_uppercase;  \
-    return MGP_ERROR_NO_ERROR;                                           \
+    return mgp_error::MGP_ERROR_NO_ERROR;                                \
   }
 
 DEFINE_MGP_VALUE_IS(null, NULL)
@@ -874,27 +874,27 @@ DEFINE_MGP_VALUE_IS(duration, DURATION)
 
 mgp_error mgp_value_get_bool(mgp_value *val, int *result) {
   *result = val->bool_v ? 1 : 0;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 mgp_error mgp_value_get_int(mgp_value *val, int64_t *result) {
   *result = val->int_v;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 mgp_error mgp_value_get_double(mgp_value *val, double *result) {
   *result = val->double_v;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 mgp_error mgp_value_get_string(mgp_value *val, const char **result) {
   static_assert(noexcept(val->string_v.c_str()));
   *result = val->string_v.c_str();
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 // NOLINTNEXTLINE(cppcoreguidelines-macro-usage)
 #define DEFINE_MGP_VALUE_GET(type)                                      \
   mgp_error mgp_value_get_##type(mgp_value *val, mgp_##type **result) { \
     *result = val->type##_v;                                            \
-    return MGP_ERROR_NO_ERROR;                                          \
+    return mgp_error::MGP_ERROR_NO_ERROR;                               \
   }
 
 DEFINE_MGP_VALUE_GET(list)
@@ -940,13 +940,13 @@ mgp_error mgp_list_append_extend(mgp_list *list, mgp_value *val) {
 mgp_error mgp_list_size(mgp_list *list, size_t *result) {
   static_assert(noexcept(list->elems.size()));
   *result = list->elems.size();
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_list_capacity(mgp_list *list, size_t *result) {
   static_assert(noexcept(list->elems.capacity()));
   *result = list->elems.capacity();
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_list_at(mgp_list *list, size_t i, mgp_value **result) {
@@ -978,7 +978,7 @@ mgp_error mgp_map_insert(mgp_map *map, const char *key, mgp_value *value) {
 mgp_error mgp_map_size(mgp_map *map, size_t *result) {
   static_assert(noexcept(map->items.size()));
   *result = map->items.size();
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_map_at(mgp_map *map, const char *key, mgp_value **result) {
@@ -1089,7 +1089,7 @@ size_t MgpPathSize(const mgp_path &path) noexcept { return path.edges.size(); }
 
 mgp_error mgp_path_size(mgp_path *path, size_t *result) {
   *result = MgpPathSize(*path);
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_path_vertex_at(mgp_path *path, size_t i, mgp_vertex **result) {
@@ -1690,7 +1690,7 @@ mgp_error mgp_vertex_equal(mgp_vertex *v1, mgp_vertex *v2, int *result) {
   // NOLINTNEXTLINE(clang-diagnostic-unevaluated-expression)
   static_assert(noexcept(*result = *v1 == *v2 ? 1 : 0));
   *result = *v1 == *v2 ? 1 : 0;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_vertex_labels_count(mgp_vertex *v, size_t *result) {
@@ -1950,7 +1950,7 @@ mgp_error mgp_edge_equal(mgp_edge *e1, mgp_edge *e2, int *result) {
   // NOLINTNEXTLINE(clang-diagnostic-unevaluated-expression)
   static_assert(noexcept(*result = *e1 == *e2 ? 1 : 0));
   *result = *e1 == *e2 ? 1 : 0;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_edge_get_type(mgp_edge *e, mgp_edge_type *result) {
@@ -1967,12 +1967,12 @@ mgp_error mgp_edge_get_type(mgp_edge *e, mgp_edge_type *result) {
 
 mgp_error mgp_edge_get_from(mgp_edge *e, mgp_vertex **result) {
   *result = &e->from;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_edge_get_to(mgp_edge *e, mgp_vertex **result) {
   *result = &e->to;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_edge_get_property(mgp_edge *e, const char *name, mgp_memory *memory, mgp_value **result) {
@@ -2082,7 +2082,7 @@ mgp_error mgp_graph_get_vertex_by_id(mgp_graph *graph, mgp_vertex_id id, mgp_mem
 
 mgp_error mgp_graph_is_mutable(mgp_graph *graph, int *result) {
   *result = MgpGraphIsMutable(*graph) ? 1 : 0;
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 };
 
 mgp_error mgp_graph_create_vertex(struct mgp_graph *graph, mgp_memory *memory, mgp_vertex **result) {
@@ -2507,7 +2507,7 @@ mgp_error mgp_proc_add_result(mgp_proc *proc, const char *name, mgp_type *type)
 
 mgp_error MgpTransAddFixedResult(mgp_trans *trans) noexcept {
   if (const auto err = AddResultToProp(trans, "query", Call<mgp_type *>(mgp_type_string), false);
-      err != MGP_ERROR_NO_ERROR) {
+      err != mgp_error::MGP_ERROR_NO_ERROR) {
     return err;
   }
   return AddResultToProp(trans, "parameters", Call<mgp_type *>(mgp_type_nullable, Call<mgp_type *>(mgp_type_map)),
@@ -2754,7 +2754,7 @@ mgp_error mgp_message_offset(struct mgp_message *message, int64_t *result) {
 mgp_error mgp_messages_size(mgp_messages *messages, size_t *result) {
   static_assert(noexcept(messages->messages.size()));
   *result = messages->messages.size();
-  return MGP_ERROR_NO_ERROR;
+  return mgp_error::MGP_ERROR_NO_ERROR;
 }
 
 mgp_error mgp_messages_at(mgp_messages *messages, size_t index, mgp_message **result) {

src/query/procedure/module.cpp

@@ -121,18 +121,18 @@ void RegisterMgLoad(ModuleRegistry *module_registry, utils::RWLock *lock, Builti
     bool succ = false;
     WithUpgradedLock(lock, [&]() {
       const char *arg_as_string{nullptr};
-      if (const auto err = mgp_value_get_string(arg, &arg_as_string); err != MGP_ERROR_NO_ERROR) {
+      if (const auto err = mgp_value_get_string(arg, &arg_as_string); err != mgp_error::MGP_ERROR_NO_ERROR) {
         succ = false;
       } else {
         succ = module_registry->LoadOrReloadModuleFromName(arg_as_string);
       }
     });
     if (!succ) {
-      MG_ASSERT(mgp_result_set_error_msg(result, "Failed to (re)load the module.") == MGP_ERROR_NO_ERROR);
+      MG_ASSERT(mgp_result_set_error_msg(result, "Failed to (re)load the module.") == mgp_error::MGP_ERROR_NO_ERROR);
     }
   };
   mgp_proc load("load", load_cb, utils::NewDeleteResource());
-  MG_ASSERT(mgp_proc_add_arg(&load, "module_name", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&load, "module_name", Call<mgp_type *>(mgp_type_string)) == mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("load", std::move(load));
 }
 
@@ -235,11 +235,16 @@ void RegisterMgProcedures(
     }
   };
   mgp_proc procedures("procedures", procedures_cb, utils::NewDeleteResource());
-  MG_ASSERT(mgp_proc_add_result(&procedures, "name", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&procedures, "signature", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&procedures, "is_write", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&procedures, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&procedures, "is_editable", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "name", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "signature", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "is_write", Call<mgp_type *>(mgp_type_bool)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "is_editable", Call<mgp_type *>(mgp_type_bool)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("procedures", std::move(procedures));
 }
 
@@ -298,9 +303,12 @@ void RegisterMgTransformations(const std::map<std::string, std::unique_ptr<Modul
     }
   };
   mgp_proc procedures("transformations", transformations_cb, utils::NewDeleteResource());
-  MG_ASSERT(mgp_proc_add_result(&procedures, "name", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&procedures, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&procedures, "is_editable", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "name", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&procedures, "is_editable", Call<mgp_type *>(mgp_type_bool)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("transformations", std::move(procedures));
 }
 
@@ -374,10 +382,14 @@ void RegisterMgFunctions(
     }
   };
   mgp_proc functions("functions", functions_cb, utils::NewDeleteResource());
-  MG_ASSERT(mgp_proc_add_result(&functions, "name", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&functions, "signature", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&functions, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&functions, "is_editable", Call<mgp_type *>(mgp_type_bool)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&functions, "name", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&functions, "signature", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&functions, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&functions, "is_editable", Call<mgp_type *>(mgp_type_bool)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("functions", std::move(functions));
 }
 namespace {
@@ -469,9 +481,10 @@ void RegisterMgGetModuleFiles(ModuleRegistry *module_registry, BuiltinModule *mo
 
   mgp_proc get_module_files("get_module_files", get_module_files_cb, utils::NewDeleteResource(),
                             {.required_privilege = AuthQuery::Privilege::MODULE_READ});
-  MG_ASSERT(mgp_proc_add_result(&get_module_files, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&get_module_files, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   MG_ASSERT(mgp_proc_add_result(&get_module_files, "is_editable", Call<mgp_type *>(mgp_type_bool)) ==
-            MGP_ERROR_NO_ERROR);
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("get_module_files", std::move(get_module_files));
 }
 
@@ -530,8 +543,10 @@ void RegisterMgGetModuleFile(ModuleRegistry *module_registry, BuiltinModule *mod
   };
   mgp_proc get_module_file("get_module_file", std::move(get_module_file_cb), utils::NewDeleteResource(),
                            {.required_privilege = AuthQuery::Privilege::MODULE_READ});
-  MG_ASSERT(mgp_proc_add_arg(&get_module_file, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&get_module_file, "content", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&get_module_file, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&get_module_file, "content", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("get_module_file", std::move(get_module_file));
 }
 
@@ -609,9 +624,12 @@ void RegisterMgCreateModuleFile(ModuleRegistry *module_registry, utils::RWLock *
   };
   mgp_proc create_module_file("create_module_file", std::move(create_module_file_cb), utils::NewDeleteResource(),
                               {.required_privilege = AuthQuery::Privilege::MODULE_WRITE});
-  MG_ASSERT(mgp_proc_add_arg(&create_module_file, "filename", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_arg(&create_module_file, "content", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_result(&create_module_file, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&create_module_file, "filename", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&create_module_file, "content", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_result(&create_module_file, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("create_module_file", std::move(create_module_file));
 }
 
@@ -664,8 +682,10 @@ void RegisterMgUpdateModuleFile(ModuleRegistry *module_registry, utils::RWLock *
   };
   mgp_proc update_module_file("update_module_file", std::move(update_module_file_cb), utils::NewDeleteResource(),
                               {.required_privilege = AuthQuery::Privilege::MODULE_WRITE});
-  MG_ASSERT(mgp_proc_add_arg(&update_module_file, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
-  MG_ASSERT(mgp_proc_add_arg(&update_module_file, "content", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&update_module_file, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&update_module_file, "content", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("update_module_file", std::move(update_module_file));
 }
 
@@ -721,7 +741,8 @@ void RegisterMgDeleteModuleFile(ModuleRegistry *module_registry, utils::RWLock *
   };
   mgp_proc delete_module_file("delete_module_file", std::move(delete_module_file_cb), utils::NewDeleteResource(),
                               {.required_privilege = AuthQuery::Privilege::MODULE_WRITE});
-  MG_ASSERT(mgp_proc_add_arg(&delete_module_file, "path", Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+  MG_ASSERT(mgp_proc_add_arg(&delete_module_file, "path", Call<mgp_type *>(mgp_type_string)) ==
+            mgp_error::MGP_ERROR_NO_ERROR);
   module->AddProcedure("delete_module_file", std::move(delete_module_file));
 }
 
@@ -801,7 +822,8 @@ bool SharedLibraryModule::Load(const std::filesystem::path &file_path) {
   spdlog::info("Loading module {}...", file_path);
   file_path_ = file_path;
   dlerror();  // Clear any existing error.
-  handle_ = dlopen(file_path.c_str(), RTLD_NOW | RTLD_LOCAL);
+  // NOLINTNEXTLINE(hicpp-signed-bitwise)
+  handle_ = dlopen(file_path.c_str(), RTLD_NOW | RTLD_LOCAL | RTLD_DEEPBIND);
   if (!handle_) {
     spdlog::error(
         utils::MessageWithLink("Unable to load module {}; {}.", file_path, dlerror(), "https://memgr.ph/modules"));
@@ -832,7 +854,7 @@ bool SharedLibraryModule::Load(const std::filesystem::path &file_path) {
       return with_error(error);
     }
     for (auto &trans : module_def->transformations) {
-      const bool success = MGP_ERROR_NO_ERROR == MgpTransAddFixedResult(&trans.second);
+      const bool success = mgp_error::MGP_ERROR_NO_ERROR == MgpTransAddFixedResult(&trans.second);
       if (!success) {
         const auto error =
             fmt::format("Unable to add result to transformation in module {}; add result failed", file_path);
@@ -941,7 +963,7 @@ bool PythonModule::Load(const std::filesystem::path &file_path) {
   auto module_cb = [&](auto *module_def, auto * /*memory*/) {
     auto result = ImportPyModule(file_path.stem().c_str(), module_def);
     for (auto &trans : module_def->transformations) {
-      succ = MgpTransAddFixedResult(&trans.second) == MGP_ERROR_NO_ERROR;
+      succ = MgpTransAddFixedResult(&trans.second) == mgp_error::MGP_ERROR_NO_ERROR;
       if (!succ) {
         return result;
       }
@@ -1032,7 +1054,7 @@ std::unique_ptr<Module> LoadModuleFromFile(const std::filesystem::path &path) {
 
 }  // namespace
 
-bool ModuleRegistry::RegisterModule(const std::string_view &name, std::unique_ptr<Module> module) {
+bool ModuleRegistry::RegisterModule(const std::string_view name, std::unique_ptr<Module> module) {
   MG_ASSERT(!name.empty(), "Module name cannot be empty");
   MG_ASSERT(module, "Tried to register an invalid module");
   if (modules_.find(name) != modules_.end()) {
@@ -1141,7 +1163,7 @@ void ModuleRegistry::UnloadAndLoadModulesFromDirectories() {
   }
 }
 
-ModulePtr ModuleRegistry::GetModuleNamed(const std::string_view &name) const {
+ModulePtr ModuleRegistry::GetModuleNamed(const std::string_view name) const {
   std::shared_lock<utils::RWLock> guard(lock_);
   auto found_it = modules_.find(name);
   if (found_it == modules_.end()) return nullptr;

src/query/procedure/module.hpp

@@ -13,6 +13,7 @@
 /// API for loading and registering modules providing custom oC procedures
 #pragma once
 
+#include <dlfcn.h>
 #include <filesystem>
 #include <functional>
 #include <optional>
@@ -76,7 +77,7 @@ class ModuleRegistry final {
   mutable utils::RWLock lock_{utils::RWLock::Priority::WRITE};
   std::unique_ptr<utils::MemoryResource> shared_{std::make_unique<utils::ResourceWithOutOfMemoryException>()};
 
-  bool RegisterModule(const std::string_view &name, std::unique_ptr<Module> module);
+  bool RegisterModule(std::string_view name, std::unique_ptr<Module> module);
 
   void DoUnloadAllModules();
 
@@ -104,7 +105,7 @@ class ModuleRegistry final {
   ///
   /// Return true if the module was loaded or reloaded successfully, false
   /// otherwise.
-  bool LoadOrReloadModuleFromName(const std::string_view name);
+  bool LoadOrReloadModuleFromName(std::string_view name);
 
   /// Atomically unload all modules and then load all possible modules from the
   /// set directories.
@@ -114,7 +115,7 @@ class ModuleRegistry final {
 
   /// Find a module with given name or return nullptr.
   /// Takes a read lock.
-  ModulePtr GetModuleNamed(const std::string_view &name) const;
+  ModulePtr GetModuleNamed(std::string_view name) const;
 
   /// Remove all loaded (non-builtin) modules.
   /// Takes a write lock.
@@ -128,6 +129,40 @@ class ModuleRegistry final {
   const std::filesystem::path &InternalModuleDir() const noexcept;
 
  private:
+  class SharedLibraryHandle {
+   public:
+    SharedLibraryHandle(const std::string &shared_library, int mode) : handle_{dlopen(shared_library.c_str(), mode)} {}
+    SharedLibraryHandle(const SharedLibraryHandle &) = delete;
+    SharedLibraryHandle(SharedLibraryHandle &&) = delete;
+    SharedLibraryHandle operator=(const SharedLibraryHandle &) = delete;
+    SharedLibraryHandle operator=(SharedLibraryHandle &&) = delete;
+
+    ~SharedLibraryHandle() {
+      if (handle_) {
+        dlclose(handle_);
+      }
+    }
+
+   private:
+    void *handle_;
+  };
+
+#if __has_feature(address_sanitizer)
+  // This is why we need RTLD_NODELETE and we must not use RTLD_DEEPBIND with
+  // ASAN: https://github.com/google/sanitizers/issues/89
+  SharedLibraryHandle libstd_handle{"libstdc++.so.6", RTLD_NOW | RTLD_LOCAL | RTLD_NODELETE};
+#else
+  // The reason behind opening share library during runtime is to avoid issues
+  // with loading symbols from stdlib. We have encounter issues with locale
+  // that cause std::cout not being printed and issues when python libraries
+  // would call stdlib (e.g. pytorch).
+  // The way that those issues were solved was
+  // by using RTLD_DEEPBIND. RTLD_DEEPBIND ensures that the lookup for the
+  // mentioned library will be first performed in the already existing binded
+  // libraries and then the global namespace.
+  // RTLD_DEEPBIND => https://linux.die.net/man/3/dlopen
+  SharedLibraryHandle libstd_handle{"libstdc++.so.6", RTLD_NOW | RTLD_LOCAL | RTLD_DEEPBIND};
+#endif
   std::vector<std::filesystem::path> modules_dirs_;
   std::filesystem::path internal_module_dir_;
 };
@@ -140,7 +175,7 @@ extern ModuleRegistry gModuleRegistry;
 /// inside this function. ModulePtr must be kept alive to make sure it won't be
 /// unloaded.
 std::optional<std::pair<procedure::ModulePtr, const mgp_proc *>> FindProcedure(
-    const ModuleRegistry &module_registry, const std::string_view fully_qualified_procedure_name,
+    const ModuleRegistry &module_registry, std::string_view fully_qualified_procedure_name,
     utils::MemoryResource *memory);
 
 /// Return the ModulePtr and `mgp_trans *` of the found transformation after resolving
@@ -148,7 +183,7 @@ std::optional<std::pair<procedure::ModulePtr, const mgp_proc *>> FindProcedure(
 /// inside this function. ModulePtr must be kept alive to make sure it won't be
 /// unloaded.
 std::optional<std::pair<procedure::ModulePtr, const mgp_trans *>> FindTransformation(
-    const ModuleRegistry &module_registry, const std::string_view fully_qualified_transformation_name,
+    const ModuleRegistry &module_registry, std::string_view fully_qualified_transformation_name,
     utils::MemoryResource *memory);
 
 /// Return the ModulePtr and `mgp_func *` of the found function after resolving
@@ -156,7 +191,7 @@ std::optional<std::pair<procedure::ModulePtr, const mgp_trans *>> FindTransforma
 /// std::nullopt is returned. `memory` is used for temporary allocations
 /// inside this function. ModulePtr must be kept alive to make sure it won't be unloaded.
 std::optional<std::pair<procedure::ModulePtr, const mgp_func *>> FindFunction(
-    const ModuleRegistry &module_registry, const std::string_view fully_qualified_function_name,
+    const ModuleRegistry &module_registry, std::string_view fully_qualified_function_name,
     utils::MemoryResource *memory);
 
 template <typename T>

src/query/procedure/py_module.cpp

@@ -55,49 +55,49 @@ PyObject *gMgpSerializationError{nullptr};       // NOLINT(cppcoreguidelines-avo
 // Returns true if an exception is raised
 bool RaiseExceptionFromErrorCode(const mgp_error error) {
   switch (error) {
-    case MGP_ERROR_NO_ERROR:
+    case mgp_error::MGP_ERROR_NO_ERROR:
       return false;
-    case MGP_ERROR_UNKNOWN_ERROR: {
+    case mgp_error::MGP_ERROR_UNKNOWN_ERROR: {
       PyErr_SetString(gMgpUnknownError, "Unknown error happened.");
       return true;
     }
-    case MGP_ERROR_UNABLE_TO_ALLOCATE: {
+    case mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE: {
       PyErr_SetString(gMgpUnableToAllocateError, "Unable to allocate memory.");
       return true;
     }
-    case MGP_ERROR_INSUFFICIENT_BUFFER: {
+    case mgp_error::MGP_ERROR_INSUFFICIENT_BUFFER: {
       PyErr_SetString(gMgpInsufficientBufferError, "Insufficient buffer.");
       return true;
     }
-    case MGP_ERROR_OUT_OF_RANGE: {
+    case mgp_error::MGP_ERROR_OUT_OF_RANGE: {
       PyErr_SetString(gMgpOutOfRangeError, "Out of range.");
       return true;
     }
-    case MGP_ERROR_LOGIC_ERROR: {
+    case mgp_error::MGP_ERROR_LOGIC_ERROR: {
       PyErr_SetString(gMgpLogicErrorError, "Logic error.");
       return true;
     }
-    case MGP_ERROR_DELETED_OBJECT: {
+    case mgp_error::MGP_ERROR_DELETED_OBJECT: {
       PyErr_SetString(gMgpDeletedObjectError, "Accessing deleted object.");
       return true;
     }
-    case MGP_ERROR_INVALID_ARGUMENT: {
+    case mgp_error::MGP_ERROR_INVALID_ARGUMENT: {
       PyErr_SetString(gMgpInvalidArgumentError, "Invalid argument.");
       return true;
     }
-    case MGP_ERROR_KEY_ALREADY_EXISTS: {
+    case mgp_error::MGP_ERROR_KEY_ALREADY_EXISTS: {
       PyErr_SetString(gMgpKeyAlreadyExistsError, "Key already exists.");
       return true;
     }
-    case MGP_ERROR_IMMUTABLE_OBJECT: {
+    case mgp_error::MGP_ERROR_IMMUTABLE_OBJECT: {
       PyErr_SetString(gMgpImmutableObjectError, "Cannot modify immutable object.");
       return true;
     }
-    case MGP_ERROR_VALUE_CONVERSION: {
+    case mgp_error::MGP_ERROR_VALUE_CONVERSION: {
       PyErr_SetString(gMgpValueConversionError, "Value conversion failed.");
       return true;
     }
-    case MGP_ERROR_SERIALIZATION_ERROR: {
+    case mgp_error::MGP_ERROR_SERIALIZATION_ERROR: {
       PyErr_SetString(gMgpSerializationError, "Operation cannot be serialized.");
       return true;
     }
@@ -902,7 +902,7 @@ std::optional<py::ExceptionInfo> AddRecordFromPython(mgp_result *result, py::Obj
     if (field_val == nullptr) {
       return py::FetchError();
     }
-    if (mgp_result_record_insert(record, field_name, field_val) != MGP_ERROR_NO_ERROR) {
+    if (mgp_result_record_insert(record, field_name, field_val) != mgp_error::MGP_ERROR_NO_ERROR) {
       std::stringstream ss;
       ss << "Unable to insert field '" << py::Object::FromBorrow(key) << "' with value: '"
          << py::Object::FromBorrow(val) << "'; did you set the correct field type?";
@@ -2281,9 +2281,10 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
   auto py_seq_to_list = [memory](PyObject *seq, Py_ssize_t len, const auto &py_seq_get_item) {
     static_assert(std::numeric_limits<Py_ssize_t>::max() <= std::numeric_limits<size_t>::max());
     MgpUniquePtr<mgp_list> list{nullptr, &mgp_list_destroy};
-    if (const auto err = CreateMgpObject(list, mgp_list_make_empty, len, memory); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = CreateMgpObject(list, mgp_list_make_empty, len, memory);
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during making mgp_list"};
     }
     for (Py_ssize_t i = 0; i < len; ++i) {
@@ -2292,17 +2293,17 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
       v = PyObjectToMgpValue(e, memory);
       const auto err = mgp_list_append(list.get(), v);
       mgp_value_destroy(v);
-      if (err != MGP_ERROR_NO_ERROR) {
-        if (err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+      if (err != mgp_error::MGP_ERROR_NO_ERROR) {
+        if (err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
           throw std::bad_alloc{};
         }
         throw std::runtime_error{"Unexpected error during appending to mgp_list"};
       }
     }
     mgp_value *v{nullptr};
-    if (const auto err = mgp_value_make_list(list.get(), &v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_list(list.get(), &v); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during making mgp_value"};
     }
     static_cast<void>(list.release());
@@ -2334,7 +2335,7 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
   };
 
   mgp_value *mgp_v{nullptr};
-  mgp_error last_error{MGP_ERROR_NO_ERROR};
+  mgp_error last_error{mgp_error::MGP_ERROR_NO_ERROR};
 
   if (o == Py_None) {
     last_error = mgp_value_make_null(memory, &mgp_v);
@@ -2360,10 +2361,10 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     MgpUniquePtr<mgp_map> map{nullptr, mgp_map_destroy};
     const auto map_err = CreateMgpObject(map, mgp_map_make_empty, memory);
 
-    if (map_err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (map_err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
     }
-    if (map_err != MGP_ERROR_NO_ERROR) {
+    if (map_err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during creating mgp_map"};
     }
 
@@ -2384,16 +2385,16 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
 
       MgpUniquePtr<mgp_value> v{PyObjectToMgpValue(value, memory), mgp_value_destroy};
 
-      if (const auto err = mgp_map_insert(map.get(), k, v.get()); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+      if (const auto err = mgp_map_insert(map.get(), k, v.get()); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
         throw std::bad_alloc{};
-      } else if (err != MGP_ERROR_NO_ERROR) {
+      } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
         throw std::runtime_error{"Unexpected error during inserting an item to mgp_map"};
       }
     }
 
-    if (const auto err = mgp_value_make_map(map.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_map(map.get(), &mgp_v); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during creating mgp_value"};
     }
     static_cast<void>(map.release());
@@ -2402,14 +2403,14 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     // Copy the edge and pass the ownership to the created mgp_value.
 
     if (const auto err = CreateMgpObject(e, mgp_edge_copy, reinterpret_cast<PyEdge *>(o)->edge, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during copying mgp_edge"};
     }
-    if (const auto err = mgp_value_make_edge(e.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_edge(e.get(), &mgp_v); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during copying mgp_edge"};
     }
     static_cast<void>(e.release());
@@ -2418,14 +2419,14 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     // Copy the edge and pass the ownership to the created mgp_value.
 
     if (const auto err = CreateMgpObject(p, mgp_path_copy, reinterpret_cast<PyPath *>(o)->path, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during copying mgp_path"};
     }
-    if (const auto err = mgp_value_make_path(p.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_path(p.get(), &mgp_v); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during copying mgp_path"};
     }
     static_cast<void>(p.release());
@@ -2434,14 +2435,14 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     // Copy the edge and pass the ownership to the created mgp_value.
 
     if (const auto err = CreateMgpObject(v, mgp_vertex_copy, reinterpret_cast<PyVertex *>(o)->vertex, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during copying mgp_vertex"};
     }
-    if (const auto err = mgp_value_make_vertex(v.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_vertex(v.get(), &mgp_v); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error during copying mgp_vertex"};
     }
     static_cast<void>(v.release());
@@ -2474,14 +2475,14 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     MgpUniquePtr<mgp_date> date{nullptr, mgp_date_destroy};
 
     if (const auto err = CreateMgpObject(date, mgp_date_from_parameters, &parameters, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_date"};
     }
-    if (const auto err = mgp_value_make_date(date.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_date(date.get(), &mgp_v); err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_value"};
     }
     static_cast<void>(date.release());
@@ -2499,14 +2500,15 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     MgpUniquePtr<mgp_local_time> local_time{nullptr, mgp_local_time_destroy};
 
     if (const auto err = CreateMgpObject(local_time, mgp_local_time_from_parameters, &parameters, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_local_time"};
     }
-    if (const auto err = mgp_value_make_local_time(local_time.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_local_time(local_time.get(), &mgp_v);
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_value"};
     }
     static_cast<void>(local_time.release());
@@ -2531,15 +2533,15 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     MgpUniquePtr<mgp_local_date_time> local_date_time{nullptr, mgp_local_date_time_destroy};
 
     if (const auto err = CreateMgpObject(local_date_time, mgp_local_date_time_from_parameters, &parameters, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_local_date_time"};
     }
     if (const auto err = mgp_value_make_local_date_time(local_date_time.get(), &mgp_v);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_value"};
     }
     static_cast<void>(local_date_time.release());
@@ -2558,14 +2560,15 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     MgpUniquePtr<mgp_duration> duration{nullptr, mgp_duration_destroy};
 
     if (const auto err = CreateMgpObject(duration, mgp_duration_from_microseconds, microseconds, memory);
-        err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_duration"};
     }
-    if (const auto err = mgp_value_make_duration(duration.get(), &mgp_v); err == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+    if (const auto err = mgp_value_make_duration(duration.get(), &mgp_v);
+        err == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
       throw std::bad_alloc{};
-    } else if (err != MGP_ERROR_NO_ERROR) {
+    } else if (err != mgp_error::MGP_ERROR_NO_ERROR) {
       throw std::runtime_error{"Unexpected error while creating mgp_value"};
     }
     static_cast<void>(duration.release());
@@ -2573,10 +2576,10 @@ mgp_value *PyObjectToMgpValue(PyObject *o, mgp_memory *memory) {
     throw std::invalid_argument("Unsupported PyObject conversion");
   }
 
-  if (last_error == MGP_ERROR_UNABLE_TO_ALLOCATE) {
+  if (last_error == mgp_error::MGP_ERROR_UNABLE_TO_ALLOCATE) {
     throw std::bad_alloc{};
   }
-  if (last_error != MGP_ERROR_NO_ERROR) {
+  if (last_error != mgp_error::MGP_ERROR_NO_ERROR) {
     throw std::runtime_error{"Unexpected error while creating mgp_value"};
   }
 

src/query/stream/common.hpp

@@ -52,10 +52,11 @@ concept Stream = requires(TStream stream) {
   typename TStream::Message;
   TStream{std::string{""}, typename TStream::StreamInfo{}, ConsumerFunction<typename TStream::Message>{}};
   { stream.Start() } -> std::same_as<void>;
+  { stream.StartWithLimit(uint64_t{}, std::optional<std::chrono::milliseconds>{}) } -> std::same_as<void>;
   { stream.Stop() } -> std::same_as<void>;
   { stream.IsRunning() } -> std::same_as<bool>;
   {
-    stream.Check(std::optional<std::chrono::milliseconds>{}, std::optional<int64_t>{},
+    stream.Check(std::optional<std::chrono::milliseconds>{}, std::optional<uint64_t>{},
                  ConsumerFunction<typename TStream::Message>{})
     } -> std::same_as<void>;
   requires std::same_as<std::decay_t<decltype(std::declval<typename TStream::StreamInfo>().common_info)>,

src/query/stream/sources.cpp

@@ -44,10 +44,13 @@ KafkaStream::StreamInfo KafkaStream::Info(std::string transformation_name) const
 }
 
 void KafkaStream::Start() { consumer_->Start(); }
+void KafkaStream::StartWithLimit(uint64_t batch_limit, std::optional<std::chrono::milliseconds> timeout) const {
+  consumer_->StartWithLimit(batch_limit, timeout);
+}
 void KafkaStream::Stop() { consumer_->Stop(); }
 bool KafkaStream::IsRunning() const { return consumer_->IsRunning(); }
 
-void KafkaStream::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+void KafkaStream::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> batch_limit,
                         const ConsumerFunction<integrations::kafka::Message> &consumer_function) const {
   consumer_->Check(timeout, batch_limit, consumer_function);
 }
@@ -106,10 +109,12 @@ PulsarStream::StreamInfo PulsarStream::Info(std::string transformation_name) con
 }
 
 void PulsarStream::Start() { consumer_->Start(); }
+void PulsarStream::StartWithLimit(uint64_t batch_limit, std::optional<std::chrono::milliseconds> timeout) const {
+  consumer_->StartWithLimit(batch_limit, timeout);
+}
 void PulsarStream::Stop() { consumer_->Stop(); }
 bool PulsarStream::IsRunning() const { return consumer_->IsRunning(); }
-
-void PulsarStream::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+void PulsarStream::Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> batch_limit,
                          const ConsumerFunction<Message> &consumer_function) const {
   consumer_->Check(timeout, batch_limit, consumer_function);
 }

src/query/stream/sources.hpp

@@ -36,10 +36,11 @@ struct KafkaStream {
   StreamInfo Info(std::string transformation_name) const;
 
   void Start();
+  void StartWithLimit(uint64_t batch_limit, std::optional<std::chrono::milliseconds> timeout) const;
   void Stop();
   bool IsRunning() const;
 
-  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> batch_limit,
              const ConsumerFunction<Message> &consumer_function) const;
 
   utils::BasicResult<std::string> SetStreamOffset(int64_t offset);
@@ -71,10 +72,11 @@ struct PulsarStream {
   StreamInfo Info(std::string transformation_name) const;
 
   void Start();
+  void StartWithLimit(uint64_t batch_limit, std::optional<std::chrono::milliseconds> timeout) const;
   void Stop();
   bool IsRunning() const;
 
-  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<int64_t> batch_limit,
+  void Check(std::optional<std::chrono::milliseconds> timeout, std::optional<uint64_t> batch_limit,
              const ConsumerFunction<Message> &consumer_function) const;
 
  private:

src/query/stream/streams.cpp

@@ -43,6 +43,7 @@ extern const Event MessagesConsumed;
 namespace memgraph::query::stream {
 namespace {
 inline constexpr auto kExpectedTransformationResultSize = 2;
+inline constexpr auto kCheckStreamResultSize = 2;
 const utils::pmr::string query_param_name{"query", utils::NewDeleteResource()};
 const utils::pmr::string params_param_name{"parameters", utils::NewDeleteResource()};
 
@@ -181,25 +182,27 @@ void Streams::RegisterKafkaProcedures() {
       const auto offset = procedure::Call<int64_t>(mgp_value_get_int, arg_offset);
       auto lock_ptr = streams_.Lock();
       auto it = GetStream(*lock_ptr, std::string(stream_name));
-      std::visit(utils::Overloaded{
-                     [&](StreamData<KafkaStream> &kafka_stream) {
-                       auto stream_source_ptr = kafka_stream.stream_source->Lock();
-                       const auto error = stream_source_ptr->SetStreamOffset(offset);
-                       if (error.HasError()) {
-                         MG_ASSERT(mgp_result_set_error_msg(result, error.GetError().c_str()) == MGP_ERROR_NO_ERROR,
-                                   "Unable to set procedure error message of procedure: {}", proc_name);
-                       }
-                     },
-                     [](auto && /*other*/) {
-                       throw QueryRuntimeException("'{}' can be only used for Kafka stream sources", proc_name);
-                     }},
+      std::visit(utils::Overloaded{[&](StreamData<KafkaStream> &kafka_stream) {
+                                     auto stream_source_ptr = kafka_stream.stream_source->Lock();
+                                     const auto error = stream_source_ptr->SetStreamOffset(offset);
+                                     if (error.HasError()) {
+                                       MG_ASSERT(mgp_result_set_error_msg(result, error.GetError().c_str()) ==
+                                                     mgp_error::MGP_ERROR_NO_ERROR,
+                                                 "Unable to set procedure error message of procedure: {}", proc_name);
+                                     }
+                                   },
+                                   [](auto && /*other*/) {
+                                     throw QueryRuntimeException("'{}' can be only used for Kafka stream sources",
+                                                                 proc_name);
+                                   }},
                  it->second);
     };
 
     mgp_proc proc(proc_name, set_stream_offset, utils::NewDeleteResource());
     MG_ASSERT(mgp_proc_add_arg(&proc, "stream_name", procedure::Call<mgp_type *>(mgp_type_string)) ==
-              MGP_ERROR_NO_ERROR);
-    MG_ASSERT(mgp_proc_add_arg(&proc, "offset", procedure::Call<mgp_type *>(mgp_type_int)) == MGP_ERROR_NO_ERROR);
+              mgp_error::MGP_ERROR_NO_ERROR);
+    MG_ASSERT(mgp_proc_add_arg(&proc, "offset", procedure::Call<mgp_type *>(mgp_type_int)) ==
+              mgp_error::MGP_ERROR_NO_ERROR);
 
     procedure::gModuleRegistry.RegisterMgProcedure(proc_name, std::move(proc));
   }
@@ -345,19 +348,19 @@ void Streams::RegisterKafkaProcedures() {
 
     mgp_proc proc(proc_name, get_stream_info, utils::NewDeleteResource());
     MG_ASSERT(mgp_proc_add_arg(&proc, "stream_name", procedure::Call<mgp_type *>(mgp_type_string)) ==
-              MGP_ERROR_NO_ERROR);
+              mgp_error::MGP_ERROR_NO_ERROR);
     MG_ASSERT(mgp_proc_add_result(&proc, consumer_group_result_name.data(),
-                                  procedure::Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+                                  procedure::Call<mgp_type *>(mgp_type_string)) == mgp_error::MGP_ERROR_NO_ERROR);
     MG_ASSERT(
         mgp_proc_add_result(&proc, topics_result_name.data(),
                             procedure::Call<mgp_type *>(mgp_type_list, procedure::Call<mgp_type *>(mgp_type_string))) ==
-        MGP_ERROR_NO_ERROR);
+        mgp_error::MGP_ERROR_NO_ERROR);
     MG_ASSERT(mgp_proc_add_result(&proc, bootstrap_servers_result_name.data(),
-                                  procedure::Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+                                  procedure::Call<mgp_type *>(mgp_type_string)) == mgp_error::MGP_ERROR_NO_ERROR);
     MG_ASSERT(mgp_proc_add_result(&proc, configs_result_name.data(), procedure::Call<mgp_type *>(mgp_type_map)) ==
-              MGP_ERROR_NO_ERROR);
+              mgp_error::MGP_ERROR_NO_ERROR);
     MG_ASSERT(mgp_proc_add_result(&proc, credentials_result_name.data(), procedure::Call<mgp_type *>(mgp_type_map)) ==
-              MGP_ERROR_NO_ERROR);
+              mgp_error::MGP_ERROR_NO_ERROR);
 
     procedure::gModuleRegistry.RegisterMgProcedure(proc_name, std::move(proc));
   }
@@ -432,14 +435,14 @@ void Streams::RegisterPulsarProcedures() {
 
     mgp_proc proc(proc_name, get_stream_info, utils::NewDeleteResource());
     MG_ASSERT(mgp_proc_add_arg(&proc, "stream_name", procedure::Call<mgp_type *>(mgp_type_string)) ==
-              MGP_ERROR_NO_ERROR);
+              mgp_error::MGP_ERROR_NO_ERROR);
     MG_ASSERT(mgp_proc_add_result(&proc, service_url_result_name.data(),
-                                  procedure::Call<mgp_type *>(mgp_type_string)) == MGP_ERROR_NO_ERROR);
+                                  procedure::Call<mgp_type *>(mgp_type_string)) == mgp_error::MGP_ERROR_NO_ERROR);
 
     MG_ASSERT(
         mgp_proc_add_result(&proc, topics_result_name.data(),
                             procedure::Call<mgp_type *>(mgp_type_list, procedure::Call<mgp_type *>(mgp_type_string))) ==
-        MGP_ERROR_NO_ERROR);
+        mgp_error::MGP_ERROR_NO_ERROR);
 
     procedure::gModuleRegistry.RegisterMgProcedure(proc_name, std::move(proc));
   }
@@ -453,7 +456,7 @@ void Streams::Create(const std::string &stream_name, typename TStream::StreamInf
 
   try {
     std::visit(
-        [&](auto &&stream_data) {
+        [&](const auto &stream_data) {
           const auto stream_source_ptr = stream_data.stream_source->ReadLock();
           Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *stream_source_ptr));
         },
@@ -572,7 +575,7 @@ void Streams::RestoreStreams() {
         auto it = CreateConsumer<T>(*locked_streams_map, stream_name, std::move(status.info), std::move(status.owner));
         if (status.is_running) {
           std::visit(
-              [&](auto &&stream_data) {
+              [&](const auto &stream_data) {
                 auto stream_source_ptr = stream_data.stream_source->Lock();
                 stream_source_ptr->Start();
               },
@@ -614,12 +617,12 @@ void Streams::Drop(const std::string &stream_name) {
   // function can be executing with the consumer, nothing else.
   // By acquiring the write lock here for the consumer, we make sure there is
   // no running Test function for this consumer, therefore it can be erased.
-  std::visit([&](auto &&stream_data) { stream_data.stream_source->Lock(); }, it->second);
+  std::visit([&](const auto &stream_data) { stream_data.stream_source->Lock(); }, it->second);
 
-  locked_streams->erase(it);
   if (!storage_.Delete(stream_name)) {
     throw StreamsException("Couldn't delete stream '{}' from persistent store!", stream_name);
   }
+  locked_streams->erase(it);
 
   // TODO(antaljanosbenjamin) Release the transformation
 }
@@ -629,7 +632,7 @@ void Streams::Start(const std::string &stream_name) {
   auto it = GetStream(*locked_streams, stream_name);
 
   std::visit(
-      [&, this](auto &&stream_data) {
+      [&, this](const auto &stream_data) {
         auto stream_source_ptr = stream_data.stream_source->Lock();
         stream_source_ptr->Start();
         Persist(CreateStatus(stream_name, stream_data.transformation_name, stream_data.owner, *stream_source_ptr));
@@ -637,12 +640,27 @@ void Streams::Start(const std::string &stream_name) {
       it->second);
 }
 
+void Streams::StartWithLimit(const std::string &stream_name, uint64_t batch_limit,
+                             std::optional<std::chrono::milliseconds> timeout) const {
+  std::optional locked_streams{streams_.ReadLock()};
+  auto it = GetStream(**locked_streams, stream_name);
+
+  std::visit(
+      [&](const auto &stream_data) {
+        const auto locked_stream_source = stream_data.stream_source->ReadLock();
+        locked_streams.reset();
+
+        locked_stream_source->StartWithLimit(batch_limit, timeout);
+      },
+      it->second);
+}
+
 void Streams::Stop(const std::string &stream_name) {
   auto locked_streams = streams_.Lock();
   auto it = GetStream(*locked_streams, stream_name);
 
   std::visit(
-      [&, this](auto &&stream_data) {
+      [&, this](const auto &stream_data) {
         auto stream_source_ptr = stream_data.stream_source->Lock();
         stream_source_ptr->Stop();
 
@@ -654,7 +672,7 @@ void Streams::Stop(const std::string &stream_name) {
 void Streams::StartAll() {
   for (auto locked_streams = streams_.Lock(); auto &[stream_name, stream_data] : *locked_streams) {
     std::visit(
-        [&stream_name = stream_name, this](auto &&stream_data) {
+        [&stream_name = stream_name, this](const auto &stream_data) {
           auto locked_stream_source = stream_data.stream_source->Lock();
           if (!locked_stream_source->IsRunning()) {
             locked_stream_source->Start();
@@ -669,7 +687,7 @@ void Streams::StartAll() {
 void Streams::StopAll() {
   for (auto locked_streams = streams_.Lock(); auto &[stream_name, stream_data] : *locked_streams) {
     std::visit(
-        [&stream_name = stream_name, this](auto &&stream_data) {
+        [&stream_name = stream_name, this](const auto &stream_data) {
           auto locked_stream_source = stream_data.stream_source->Lock();
           if (locked_stream_source->IsRunning()) {
             locked_stream_source->Stop();
@@ -686,7 +704,7 @@ std::vector<StreamStatus<>> Streams::GetStreamInfo() const {
   {
     for (auto locked_streams = streams_.ReadLock(); const auto &[stream_name, stream_data] : *locked_streams) {
       std::visit(
-          [&, &stream_name = stream_name](auto &&stream_data) {
+          [&, &stream_name = stream_name](const auto &stream_data) {
             auto locked_stream_source = stream_data.stream_source->ReadLock();
             auto info = locked_stream_source->Info(stream_data.transformation_name);
             result.emplace_back(StreamStatus<>{stream_name, StreamType(*locked_stream_source),
@@ -700,12 +718,12 @@ std::vector<StreamStatus<>> Streams::GetStreamInfo() const {
 }
 
 TransformationResult Streams::Check(const std::string &stream_name, std::optional<std::chrono::milliseconds> timeout,
-                                    std::optional<int64_t> batch_limit) const {
+                                    std::optional<uint64_t> batch_limit) const {
   std::optional locked_streams{streams_.ReadLock()};
   auto it = GetStream(**locked_streams, stream_name);
 
   return std::visit(
-      [&](auto &&stream_data) {
+      [&](const auto &stream_data) {
         // This depends on the fact that Drop will first acquire a write lock to the consumer, and erase it only after
         // that
         const auto locked_stream_source = stream_data.stream_source->ReadLock();
@@ -722,15 +740,27 @@ TransformationResult Streams::Check(const std::string &stream_name, std::optiona
           auto accessor = interpreter_context->db->Access();
           CallCustomTransformation(transformation_name, messages, result, accessor, *memory_resource, stream_name);
 
-          for (auto &row : result.rows) {
-            auto [query, parameters] = ExtractTransformationResult(row.values, transformation_name, stream_name);
-            std::vector<TypedValue> result_row;
-            result_row.reserve(kExpectedTransformationResultSize);
-            result_row.push_back(std::move(query));
-            result_row.push_back(std::move(parameters));
+          auto result_row = std::vector<TypedValue>();
+          result_row.reserve(kCheckStreamResultSize);
 
-            test_result.push_back(std::move(result_row));
-          }
+          auto queries_and_parameters = std::vector<TypedValue>(result.rows.size());
+          std::transform(
+              result.rows.cbegin(), result.rows.cend(), queries_and_parameters.begin(), [&](const auto &row) {
+                auto [query, parameters] = ExtractTransformationResult(row.values, transformation_name, stream_name);
+
+                return std::map<std::string, TypedValue>{{"query", std::move(query)},
+                                                         {"parameters", std::move(parameters)}};
+              });
+          result_row.emplace_back(std::move(queries_and_parameters));
+
+          auto messages_list = std::vector<TypedValue>(messages.size());
+          std::transform(messages.cbegin(), messages.cend(), messages_list.begin(), [](const auto &message) {
+            return std::string_view(message.Payload().data(), message.Payload().size());
+          });
+
+          result_row.emplace_back(std::move(messages_list));
+
+          test_result.emplace_back(std::move(result_row));
         };
 
         locked_stream_source->Check(timeout, batch_limit, consumer_function);

src/query/stream/streams.hpp

@@ -115,6 +115,17 @@ class Streams final {
   /// @throws ConsumerRunningException if the consumer is already running
   void Start(const std::string &stream_name);
 
+  /// Start consuming from a stream.
+  ///
+  /// @param stream_name name of the stream that needs to be started
+  /// @param batch_limit number of batches we want to consume before stopping
+  /// @param timeout the maximum duration during which the command should run.
+  ///
+  /// @throws StreamsException if the stream doesn't exist
+  /// @throws ConsumerRunningException if the consumer is already running
+  void StartWithLimit(const std::string &stream_name, uint64_t batch_limit,
+                      std::optional<std::chrono::milliseconds> timeout) const;
+
   /// Stop consuming from a stream.
   ///
   /// @param stream_name name of the stream that needs to be stopped
@@ -142,6 +153,7 @@ class Streams final {
   ///
   /// @param stream_name name of the stream we want to test
   /// @param batch_limit number of batches we want to test before stopping
+  /// @param timeout the maximum duration during which the command should run.
   ///
   /// @returns A vector of vectors of TypedValue. Each subvector contains two elements, the query string and the
   /// nullable parameters map.
@@ -151,7 +163,7 @@ class Streams final {
   /// @throws ConsumerCheckFailedException if the transformation function throws any std::exception during processing
   TransformationResult Check(const std::string &stream_name,
                              std::optional<std::chrono::milliseconds> timeout = std::nullopt,
-                             std::optional<int64_t> batch_limit = std::nullopt) const;
+                             std::optional<uint64_t> batch_limit = std::nullopt) const;
 
  private:
   template <Stream TStream>
@@ -176,7 +188,7 @@ class Streams final {
   void Persist(StreamStatus<TStream> &&status) {
     const std::string stream_name = status.name;
     if (!storage_.Put(stream_name, nlohmann::json(std::move(status)).dump())) {
-      throw StreamsException{"Couldn't persist steam data for stream '{}'", stream_name};
+      throw StreamsException{"Couldn't persist stream data for stream '{}'", stream_name};
     }
   }
 

src/query/typed_value.cpp

@@ -407,7 +407,7 @@ DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(int, Int, int_v)
 DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(bool, Bool, bool_v)
 DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(int64_t, Int, int_v)
 DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(double, Double, double_v)
-DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(const std::string_view &, String, string_v)
+DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(const std::string_view, String, string_v)
 DEFINE_TYPED_VALUE_COPY_ASSIGNMENT(const TypedValue::TVector &, List, list_v)
 
 TypedValue &TypedValue::operator=(const std::vector<TypedValue> &other) {

src/query/typed_value.hpp

@@ -185,7 +185,7 @@ class TypedValue {
     new (&string_v) TString(value, memory_);
   }
 
-  explicit TypedValue(const std::string_view &value, utils::MemoryResource *memory = utils::NewDeleteResource())
+  explicit TypedValue(const std::string_view value, utils::MemoryResource *memory = utils::NewDeleteResource())
       : memory_(memory), type_(Type::String) {
     new (&string_v) TString(value, memory_);
   }
@@ -420,7 +420,7 @@ class TypedValue {
   TypedValue &operator=(bool);
   TypedValue &operator=(int64_t);
   TypedValue &operator=(double);
-  TypedValue &operator=(const std::string_view &);
+  TypedValue &operator=(std::string_view);
   TypedValue &operator=(const TVector &);
   TypedValue &operator=(const std::vector<TypedValue> &);
   TypedValue &operator=(const TMap &);

src/slk/serialization.hpp

@@ -166,7 +166,7 @@ inline void Save(const char *obj, Builder *builder) {
   builder->Save(reinterpret_cast<const uint8_t *>(obj), size);
 }
 
-inline void Save(const std::string_view &obj, Builder *builder) {
+inline void Save(const std::string_view obj, Builder *builder) {
   uint64_t size = obj.size();
   Save(size, builder);
   builder->Save(reinterpret_cast<const uint8_t *>(obj.data()), size);

src/storage/v2/CMakeLists.txt

@@ -13,7 +13,6 @@ set(storage_v2_src_files
     storage.cpp)
 
 ##### Replication #####
-
 define_add_lcp(add_lcp_storage lcp_storage_cpp_files generated_lcp_storage_files)
 
 add_lcp_storage(replication/rpc.lcp SLK_SERIALIZE)
@@ -26,10 +25,10 @@ set(storage_v2_src_files
     replication/replication_server.cpp
     replication/serialization.cpp
     replication/slk.cpp
+    replication/replication_persistence_helper.cpp
     ${lcp_storage_cpp_files})
 
 #######################
-
 find_package(gflags REQUIRED)
 find_package(Threads REQUIRED)
 

src/storage/v2/config.hpp

@@ -49,7 +49,7 @@ struct Config {
     uint64_t wal_file_flush_every_n_tx{100000};
 
     bool snapshot_on_exit{false};
-
+    bool restore_replicas_on_startup{false};
   } durability;
 
   struct Transaction {

src/storage/v2/durability/paths.hpp

@@ -22,6 +22,7 @@ static const std::string kSnapshotDirectory{"snapshots"};
 static const std::string kWalDirectory{"wal"};
 static const std::string kBackupDirectory{".backup"};
 static const std::string kLockFile{".lock"};
+static const std::string kReplicationDirectory{"replication"};
 
 // This is the prefix used for Snapshot and WAL filenames. It is a timestamp
 // format that equals to: YYYYmmddHHMMSSffffff

src/storage/v2/durability/serialization.cpp

@@ -27,7 +27,7 @@ void WriteSize(Encoder *encoder, uint64_t size) {
 }
 }  // namespace
 
-void Encoder::Initialize(const std::filesystem::path &path, const std::string_view &magic, uint64_t version) {
+void Encoder::Initialize(const std::filesystem::path &path, const std::string_view magic, uint64_t version) {
   file_.Open(path, utils::OutputFile::Mode::OVERWRITE_EXISTING);
   Write(reinterpret_cast<const uint8_t *>(magic.data()), magic.size());
   auto version_encoded = utils::HostToLittleEndian(version);
@@ -73,7 +73,7 @@ void Encoder::WriteDouble(double value) {
   Write(reinterpret_cast<const uint8_t *>(&value_uint), sizeof(value_uint));
 }
 
-void Encoder::WriteString(const std::string_view &value) {
+void Encoder::WriteString(const std::string_view value) {
   WriteMarker(Marker::TYPE_STRING);
   WriteSize(this, value.size());
   Write(reinterpret_cast<const uint8_t *>(value.data()), value.size());

src/storage/v2/durability/serialization.hpp

@@ -34,14 +34,14 @@ class BaseEncoder {
   virtual void WriteBool(bool value) = 0;
   virtual void WriteUint(uint64_t value) = 0;
   virtual void WriteDouble(double value) = 0;
-  virtual void WriteString(const std::string_view &value) = 0;
+  virtual void WriteString(std::string_view value) = 0;
   virtual void WritePropertyValue(const PropertyValue &value) = 0;
 };
 
 /// Encoder that is used to generate a snapshot/WAL.
 class Encoder final : public BaseEncoder {
  public:
-  void Initialize(const std::filesystem::path &path, const std::string_view &magic, uint64_t version);
+  void Initialize(const std::filesystem::path &path, std::string_view magic, uint64_t version);
 
   void OpenExisting(const std::filesystem::path &path);
 
@@ -54,7 +54,7 @@ class Encoder final : public BaseEncoder {
   void WriteBool(bool value) override;
   void WriteUint(uint64_t value) override;
   void WriteDouble(double value) override;
-  void WriteString(const std::string_view &value) override;
+  void WriteString(std::string_view value) override;
   void WritePropertyValue(const PropertyValue &value) override;
 
   uint64_t GetPosition();

src/storage/v2/name_id_mapper.hpp

@@ -29,8 +29,8 @@ class NameIdMapper final {
     bool operator<(const MapNameToId &other) { return name < other.name; }
     bool operator==(const MapNameToId &other) { return name == other.name; }
 
-    bool operator<(const std::string_view &other) { return name < other; }
-    bool operator==(const std::string_view &other) { return name == other; }
+    bool operator<(const std::string_view other) const { return name < other; }
+    bool operator==(const std::string_view other) const { return name == other; }
   };
 
   struct MapIdToName {
@@ -46,7 +46,7 @@ class NameIdMapper final {
 
  public:
   /// @throw std::bad_alloc if unable to insert a new mapping
-  uint64_t NameToId(const std::string_view &name) {
+  uint64_t NameToId(const std::string_view name) {
     auto name_to_id_acc = name_to_id_.access();
     auto found = name_to_id_acc.find(name);
     uint64_t id;

src/storage/v2/replication/config.hpp

@@ -10,16 +10,23 @@
 // licenses/APL.txt.
 
 #pragma once
+
+#include <chrono>
 #include <optional>
 #include <string>
 
 namespace memgraph::storage::replication {
 struct ReplicationClientConfig {
-  std::optional<double> timeout;
+  // The default delay between main checking/pinging replicas is 1s because
+  // that seems like a reasonable timeframe in which main should notice a
+  // replica is down.
+  std::chrono::seconds replica_check_frequency{1};
 
   struct SSL {
     std::string key_file = "";
     std::string cert_file = "";
+
+    friend bool operator==(const SSL &, const SSL &) = default;
   };
 
   std::optional<SSL> ssl;

src/storage/v2/replication/enums.hpp

@@ -16,4 +16,6 @@ namespace memgraph::storage::replication {
 enum class ReplicationMode : std::uint8_t { SYNC, ASYNC };
 
 enum class ReplicaState : std::uint8_t { READY, REPLICATING, RECOVERY, INVALID };
+
+enum class RegistrationMode : std::uint8_t { MUST_BE_INSTANTLY_VALID, CAN_BE_INVALID };
 }  // namespace memgraph::storage::replication

src/storage/v2/replication/replication_client.cpp

@@ -41,11 +41,43 @@ Storage::ReplicationClient::ReplicationClient(std::string name, Storage *storage
   }
 
   rpc_client_.emplace(endpoint, &*rpc_context_);
-  TryInitializeClient();
+  TryInitializeClientSync();
 
-  if (config.timeout && replica_state_ != replication::ReplicaState::INVALID) {
-    timeout_.emplace(*config.timeout);
-    timeout_dispatcher_.emplace();
+  // Help the user to get the most accurate replica state possible.
+  if (config.replica_check_frequency > std::chrono::seconds(0)) {
+    replica_checker_.Run("Replica Checker", config.replica_check_frequency, [&] { FrequentCheck(); });
+  }
+}
+
+void Storage::ReplicationClient::TryInitializeClientAsync() {
+  thread_pool_.AddTask([this] {
+    rpc_client_->Abort();
+    this->TryInitializeClientSync();
+  });
+}
+
+void Storage::ReplicationClient::FrequentCheck() {
+  const auto is_success = std::invoke([this]() {
+    try {
+      auto stream{rpc_client_->Stream<replication::FrequentHeartbeatRpc>()};
+      const auto response = stream.AwaitResponse();
+      return response.success;
+    } catch (const rpc::RpcFailedException &) {
+      return false;
+    }
+  });
+  // States: READY, REPLICATING, RECOVERY, INVALID
+  // If success && ready, replicating, recovery -> stay the same because something good is going on.
+  // If success && INVALID -> [it's possible that replica came back to life] -> TryInitializeClient.
+  // If fail -> [replica is not reachable at all] -> INVALID state.
+  // NOTE: TryInitializeClient might return nothing if there is a branching point.
+  // NOTE: The early return pattern simplified the code, but the behavior should be as explained.
+  if (!is_success) {
+    replica_state_.store(replication::ReplicaState::INVALID);
+    return;
+  }
+  if (replica_state_.load() == replication::ReplicaState::INVALID) {
+    TryInitializeClientAsync();
   }
 }
 
@@ -100,7 +132,7 @@ void Storage::ReplicationClient::InitializeClient() {
   }
 }
 
-void Storage::ReplicationClient::TryInitializeClient() {
+void Storage::ReplicationClient::TryInitializeClientSync() {
   try {
     InitializeClient();
   } catch (const rpc::RpcFailedException &) {
@@ -113,10 +145,7 @@ void Storage::ReplicationClient::TryInitializeClient() {
 
 void Storage::ReplicationClient::HandleRpcFailure() {
   spdlog::error(utils::MessageWithLink("Couldn't replicate data to {}.", name_, "https://memgr.ph/replication"));
-  thread_pool_.AddTask([this] {
-    rpc_client_->Abort();
-    this->TryInitializeClient();
-  });
+  TryInitializeClientAsync();
 }
 
 replication::SnapshotRes Storage::ReplicationClient::TransferSnapshot(const std::filesystem::path &path) {
@@ -204,41 +233,6 @@ void Storage::ReplicationClient::FinalizeTransactionReplication() {
 
   if (mode_ == replication::ReplicationMode::ASYNC) {
     thread_pool_.AddTask([this] { this->FinalizeTransactionReplicationInternal(); });
-  } else if (timeout_) {
-    MG_ASSERT(mode_ == replication::ReplicationMode::SYNC, "Only SYNC replica can have a timeout.");
-    MG_ASSERT(timeout_dispatcher_, "Timeout thread is missing");
-    timeout_dispatcher_->WaitForTaskToFinish();
-
-    timeout_dispatcher_->active = true;
-    thread_pool_.AddTask([&, this] {
-      this->FinalizeTransactionReplicationInternal();
-      std::unique_lock main_guard(timeout_dispatcher_->main_lock);
-      // TimerThread can finish waiting for timeout
-      timeout_dispatcher_->active = false;
-      // Notify the main thread
-      timeout_dispatcher_->main_cv.notify_one();
-    });
-
-    timeout_dispatcher_->StartTimeoutTask(*timeout_);
-
-    // Wait until one of the threads notifies us that they finished executing
-    // Both threads should first set the active flag to false
-    {
-      std::unique_lock main_guard(timeout_dispatcher_->main_lock);
-      timeout_dispatcher_->main_cv.wait(main_guard, [&] { return !timeout_dispatcher_->active.load(); });
-    }
-
-    // TODO (antonio2368): Document and/or polish SEMI-SYNC to ASYNC fallback.
-    if (replica_state_ == replication::ReplicaState::REPLICATING) {
-      mode_ = replication::ReplicationMode::ASYNC;
-      timeout_.reset();
-      // This can only happen if we timeouted so we are sure that
-      // Timeout task finished
-      // We need to delete timeout dispatcher AFTER the replication
-      // finished because it tries to acquire the timeout lock
-      // and acces the `active` variable`
-      thread_pool_.AddTask([this] { timeout_dispatcher_.reset(); });
-    }
   } else {
     FinalizeTransactionReplicationInternal();
   }
@@ -503,30 +497,34 @@ std::vector<Storage::ReplicationClient::RecoveryStep> Storage::ReplicationClient
   return recovery_steps;
 }
 
-////// TimeoutDispatcher //////
-void Storage::ReplicationClient::TimeoutDispatcher::WaitForTaskToFinish() {
-  // Wait for the previous timeout task to finish
-  std::unique_lock main_guard(main_lock);
-  main_cv.wait(main_guard, [&] { return finished; });
-}
+Storage::TimestampInfo Storage::ReplicationClient::GetTimestampInfo() {
+  Storage::TimestampInfo info;
+  info.current_timestamp_of_replica = 0;
+  info.current_number_of_timestamp_behind_master = 0;
 
-void Storage::ReplicationClient::TimeoutDispatcher::StartTimeoutTask(const double timeout) {
-  timeout_pool.AddTask([timeout, this] {
-    finished = false;
-    using std::chrono::steady_clock;
-    const auto timeout_duration =
-        std::chrono::duration_cast<steady_clock::duration>(std::chrono::duration<double>(timeout));
-    const auto end_time = steady_clock::now() + timeout_duration;
-    while (active && (steady_clock::now() < end_time)) {
-      std::this_thread::sleep_for(std::chrono::milliseconds(10));
+  try {
+    auto stream{rpc_client_->Stream<replication::TimestampRpc>()};
+    const auto response = stream.AwaitResponse();
+    const auto is_success = response.success;
+    if (!is_success) {
+      replica_state_.store(replication::ReplicaState::INVALID);
+      HandleRpcFailure();
     }
+    auto main_time_stamp = storage_->last_commit_timestamp_.load();
+    info.current_timestamp_of_replica = response.current_commit_timestamp;
+    info.current_number_of_timestamp_behind_master = response.current_commit_timestamp - main_time_stamp;
+  } catch (const rpc::RpcFailedException &) {
+    {
+      std::unique_lock client_guard(client_lock_);
+      replica_state_.store(replication::ReplicaState::INVALID);
+    }
+    HandleRpcFailure();  // mutex already unlocked, if the new enqueued task dispatches immediately it probably won't
+                         // block
+  }
 
-    std::unique_lock main_guard(main_lock);
-    finished = true;
-    active = false;
-    main_cv.notify_one();
-  });
+  return info;
 }
+
 ////// ReplicaStream //////
 Storage::ReplicationClient::ReplicaStream::ReplicaStream(ReplicationClient *self,
                                                          const uint64_t previous_commit_timestamp,

src/storage/v2/replication/replication_client.hpp

@@ -120,10 +120,10 @@ class Storage::ReplicationClient {
 
   auto Mode() const { return mode_; }
 
-  auto Timeout() const { return timeout_; }
-
   const auto &Endpoint() const { return rpc_client_->Endpoint(); }
 
+  Storage::TimestampInfo GetTimestampInfo();
+
  private:
   void FinalizeTransactionReplicationInternal();
 
@@ -142,46 +142,20 @@ class Storage::ReplicationClient {
 
   std::vector<RecoveryStep> GetRecoverySteps(uint64_t replica_commit, utils::FileRetainer::FileLocker *file_locker);
 
+  void FrequentCheck();
   void InitializeClient();
-
-  void TryInitializeClient();
-
+  void TryInitializeClientSync();
+  void TryInitializeClientAsync();
   void HandleRpcFailure();
 
   std::string name_;
-
   Storage *storage_;
-
   std::optional<communication::ClientContext> rpc_context_;
   std::optional<rpc::Client> rpc_client_;
 
   std::optional<ReplicaStream> replica_stream_;
   replication::ReplicationMode mode_{replication::ReplicationMode::SYNC};
 
-  // Dispatcher class for timeout tasks
-  struct TimeoutDispatcher {
-    explicit TimeoutDispatcher(){};
-
-    void WaitForTaskToFinish();
-
-    void StartTimeoutTask(double timeout);
-
-    // If the Timeout task should continue waiting
-    std::atomic<bool> active{false};
-
-    std::mutex main_lock;
-    std::condition_variable main_cv;
-
-   private:
-    // if the Timeout task finished executing
-    bool finished{true};
-
-    utils::ThreadPool timeout_pool{1};
-  };
-
-  std::optional<double> timeout_;
-  std::optional<TimeoutDispatcher> timeout_dispatcher_;
-
   utils::SpinLock client_lock_;
   // This thread pool is used for background tasks so we don't
   // block the main storage thread
@@ -198,6 +172,8 @@ class Storage::ReplicationClient {
   //    to ignore concurrency problems inside the client.
   utils::ThreadPool thread_pool_{1};
   std::atomic<replication::ReplicaState> replica_state_{replication::ReplicaState::INVALID};
+
+  utils::Scheduler replica_checker_;
 };
 
 }  // namespace memgraph::storage

src/storage/v2/replication/replication_persistence_helper.cpp

@@ -0,0 +1,83 @@
+// Copyright 2022 Memgraph Ltd.
+//
+// Use of this software is governed by the Business Source License
+// included in the file licenses/BSL.txt; by using this file, you agree to be bound by the terms of the Business Source
+// License, and you may not use this file except in compliance with the Business Source License.
+//
+// As of the Change Date specified in that file, in accordance with
+// the Business Source License, use of this software will be governed
+// by the Apache License, Version 2.0, included in the file
+// licenses/APL.txt.
+
+#include "storage/v2/replication/replication_persistence_helper.hpp"
+#include "utils/logging.hpp"
+
+namespace {
+const std::string kReplicaName = "replica_name";
+const std::string kIpAddress = "replica_ip_address";
+const std::string kPort = "replica_port";
+const std::string kSyncMode = "replica_sync_mode";
+const std::string kCheckFrequency = "replica_check_frequency";
+const std::string kSSLKeyFile = "replica_ssl_key_file";
+const std::string kSSLCertFile = "replica_ssl_cert_file";
+}  // namespace
+
+namespace memgraph::storage::replication {
+
+nlohmann::json ReplicaStatusToJSON(ReplicaStatus &&status) {
+  auto data = nlohmann::json::object();
+
+  data[kReplicaName] = std::move(status.name);
+  data[kIpAddress] = std::move(status.ip_address);
+  data[kPort] = status.port;
+  data[kSyncMode] = status.sync_mode;
+
+  data[kCheckFrequency] = status.replica_check_frequency.count();
+
+  if (status.ssl.has_value()) {
+    data[kSSLKeyFile] = std::move(status.ssl->key_file);
+    data[kSSLCertFile] = std::move(status.ssl->cert_file);
+  } else {
+    data[kSSLKeyFile] = nullptr;
+    data[kSSLCertFile] = nullptr;
+  }
+
+  return data;
+}
+
+std::optional<ReplicaStatus> JSONToReplicaStatus(nlohmann::json &&data) {
+  ReplicaStatus replica_status;
+
+  const auto get_failed_message = [](const std::string_view message, const std::string_view nested_message) {
+    return fmt::format("Failed to deserialize replica's configuration: {} : {}", message, nested_message);
+  };
+
+  try {
+    data.at(kReplicaName).get_to(replica_status.name);
+    data.at(kIpAddress).get_to(replica_status.ip_address);
+    data.at(kPort).get_to(replica_status.port);
+    data.at(kSyncMode).get_to(replica_status.sync_mode);
+
+    replica_status.replica_check_frequency = std::chrono::seconds(data.at(kCheckFrequency));
+
+    const auto &key_file = data.at(kSSLKeyFile);
+    const auto &cert_file = data.at(kSSLCertFile);
+
+    MG_ASSERT(key_file.is_null() == cert_file.is_null());
+
+    if (!key_file.is_null()) {
+      replica_status.ssl = replication::ReplicationClientConfig::SSL{};
+      data.at(kSSLKeyFile).get_to(replica_status.ssl->key_file);
+      data.at(kSSLCertFile).get_to(replica_status.ssl->cert_file);
+    }
+  } catch (const nlohmann::json::type_error &exception) {
+    spdlog::error(get_failed_message("Invalid type conversion", exception.what()));
+    return std::nullopt;
+  } catch (const nlohmann::json::out_of_range &exception) {
+    spdlog::error(get_failed_message("Non existing field", exception.what()));
+    return std::nullopt;
+  }
+
+  return replica_status;
+}
+}  // namespace memgraph::storage::replication

src/storage/v2/replication/replication_persistence_helper.hpp

@@ -0,0 +1,40 @@
+// Copyright 2022 Memgraph Ltd.
+//
+// Use of this software is governed by the Business Source License
+// included in the file licenses/BSL.txt; by using this file, you agree to be bound by the terms of the Business Source
+// License, and you may not use this file except in compliance with the Business Source License.
+//
+// As of the Change Date specified in that file, in accordance with
+// the Business Source License, use of this software will be governed
+// by the Apache License, Version 2.0, included in the file
+// licenses/APL.txt.
+
+#pragma once
+
+#include <chrono>
+#include <compare>
+#include <optional>
+#include <string>
+
+#include <json/json.hpp>
+
+#include "storage/v2/replication/config.hpp"
+#include "storage/v2/replication/enums.hpp"
+
+namespace memgraph::storage::replication {
+
+struct ReplicaStatus {
+  std::string name;
+  std::string ip_address;
+  uint16_t port;
+  ReplicationMode sync_mode;
+  std::chrono::seconds replica_check_frequency;
+  std::optional<ReplicationClientConfig::SSL> ssl;
+
+  friend bool operator==(const ReplicaStatus &, const ReplicaStatus &) = default;
+};
+
+nlohmann::json ReplicaStatusToJSON(ReplicaStatus &&status);
+
+std::optional<ReplicaStatus> JSONToReplicaStatus(nlohmann::json &&data);
+}  // namespace memgraph::storage::replication

src/storage/v2/replication/replication_server.cpp

@@ -60,6 +60,10 @@ Storage::ReplicationServer::ReplicationServer(Storage *storage, io::network::End
     spdlog::debug("Received HeartbeatRpc");
     this->HeartbeatHandler(req_reader, res_builder);
   });
+  rpc_server_->Register<replication::FrequentHeartbeatRpc>([](auto *req_reader, auto *res_builder) {
+    spdlog::debug("Received FrequentHeartbeatRpc");
+    FrequentHeartbeatHandler(req_reader, res_builder);
+  });
   rpc_server_->Register<replication::AppendDeltasRpc>([this](auto *req_reader, auto *res_builder) {
     spdlog::debug("Received AppendDeltasRpc");
     this->AppendDeltasHandler(req_reader, res_builder);
@@ -76,6 +80,10 @@ Storage::ReplicationServer::ReplicationServer(Storage *storage, io::network::End
     spdlog::debug("Received CurrentWalRpc");
     this->CurrentWalHandler(req_reader, res_builder);
   });
+  rpc_server_->Register<replication::TimestampRpc>([this](auto *req_reader, auto *res_builder) {
+    spdlog::debug("Received TimestampRpc");
+    this->TimestampHandler(req_reader, res_builder);
+  });
   rpc_server_->Start();
 }
 
@@ -86,6 +94,13 @@ void Storage::ReplicationServer::HeartbeatHandler(slk::Reader *req_reader, slk::
   slk::Save(res, res_builder);
 }
 
+void Storage::ReplicationServer::FrequentHeartbeatHandler(slk::Reader *req_reader, slk::Builder *res_builder) {
+  replication::FrequentHeartbeatReq req;
+  slk::Load(&req, req_reader);
+  replication::FrequentHeartbeatRes res{true};
+  slk::Save(res, res_builder);
+}
+
 void Storage::ReplicationServer::AppendDeltasHandler(slk::Reader *req_reader, slk::Builder *res_builder) {
   replication::AppendDeltasReq req;
   slk::Load(&req, req_reader);
@@ -273,6 +288,14 @@ void Storage::ReplicationServer::LoadWal(replication::Decoder *decoder) {
   }
 }
 
+void Storage::ReplicationServer::TimestampHandler(slk::Reader *req_reader, slk::Builder *res_builder) {
+  replication::TimestampReq req;
+  slk::Load(&req, req_reader);
+
+  replication::TimestampRes res{true, storage_->last_commit_timestamp_.load()};
+  slk::Save(res, res_builder);
+}
+
 Storage::ReplicationServer::~ReplicationServer() {
   if (rpc_server_) {
     rpc_server_->Shutdown();

src/storage/v2/replication/replication_server.hpp

@@ -29,10 +29,12 @@ class Storage::ReplicationServer {
  private:
   // RPC handlers
   void HeartbeatHandler(slk::Reader *req_reader, slk::Builder *res_builder);
+  static void FrequentHeartbeatHandler(slk::Reader *req_reader, slk::Builder *res_builder);
   void AppendDeltasHandler(slk::Reader *req_reader, slk::Builder *res_builder);
   void SnapshotHandler(slk::Reader *req_reader, slk::Builder *res_builder);
   void WalFilesHandler(slk::Reader *req_reader, slk::Builder *res_builder);
   void CurrentWalHandler(slk::Reader *req_reader, slk::Builder *res_builder);
+  void TimestampHandler(slk::Reader *req_reader, slk::Builder *res_builder);
 
   void LoadWal(replication::Decoder *decoder);
   uint64_t ReadAndApplyDelta(durability::BaseDecoder *decoder);

src/storage/v2/replication/rpc.lcp

@@ -43,6 +43,12 @@ cpp<#
      (current-commit-timestamp :uint64_t)
      (epoch-id "std::string"))))
 
+;; FrequentHearthbeat is required because calling Heartbeat takes the storage lock.
+;; Configured by `replication_replica_check_delay`.
+(lcp:define-rpc frequent-heartbeat
+  (:request ())
+  (:response ((success :bool))))
+
 (lcp:define-rpc snapshot
   (:request ())
   (:response
@@ -61,6 +67,12 @@ cpp<#
     ((success :bool)
      (current-commit-timestamp :uint64_t))))
 
+(lcp:define-rpc timestamp
+  (:request ())
+  (:response
+    ((success :bool)
+     (current-commit-timestamp :uint64_t))))
+
 (lcp:pop-namespace) ;; replication
 (lcp:pop-namespace) ;; storage
 (lcp:pop-namespace) ;; memgraph

src/storage/v2/replication/serialization.cpp

@@ -30,7 +30,7 @@ void Encoder::WriteDouble(double value) {
   slk::Save(value, builder_);
 }
 
-void Encoder::WriteString(const std::string_view &value) {
+void Encoder::WriteString(const std::string_view value) {
   WriteMarker(durability::Marker::TYPE_STRING);
   slk::Save(value, builder_);
 }

src/storage/v2/replication/serialization.hpp

@@ -33,7 +33,7 @@ class Encoder final : public durability::BaseEncoder {
 
   void WriteDouble(double value) override;
 
-  void WriteString(const std::string_view &value) override;
+  void WriteString(std::string_view value) override;
 
   void WritePropertyValue(const PropertyValue &value) override;
 

src/storage/v2/storage.cpp

@@ -17,6 +17,7 @@
 #include <variant>
 
 #include <gflags/gflags.h>
+#include <spdlog/spdlog.h>
 
 #include "io/network/endpoint.hpp"
 #include "storage/v2/durability/durability.hpp"
@@ -28,6 +29,8 @@
 #include "storage/v2/indices.hpp"
 #include "storage/v2/mvcc.hpp"
 #include "storage/v2/replication/config.hpp"
+#include "storage/v2/replication/enums.hpp"
+#include "storage/v2/replication/replication_persistence_helper.hpp"
 #include "storage/v2/transaction.hpp"
 #include "storage/v2/vertex_accessor.hpp"
 #include "utils/file.hpp"
@@ -50,6 +53,19 @@ using OOMExceptionEnabler = utils::MemoryTracker::OutOfMemoryExceptionEnabler;
 
 namespace {
 inline constexpr uint16_t kEpochHistoryRetention = 1000;
+
+std::string RegisterReplicaErrorToString(Storage::RegisterReplicaError error) {
+  switch (error) {
+    case Storage::RegisterReplicaError::NAME_EXISTS:
+      return "NAME_EXISTS";
+    case Storage::RegisterReplicaError::END_POINT_EXISTS:
+      return "END_POINT_EXISTS";
+    case Storage::RegisterReplicaError::CONNECTION_FAILED:
+      return "CONNECTION_FAILED";
+    case Storage::RegisterReplicaError::COULD_NOT_BE_PERSISTED:
+      return "COULD_NOT_BE_PERSISTED";
+  }
+}
 }  // namespace
 
 auto AdvanceToVisibleVertex(utils::SkipList<Vertex>::Iterator it, utils::SkipList<Vertex>::Iterator end,
@@ -306,6 +322,13 @@ Storage::Storage(Config config)
       uuid_(utils::GenerateUUID()),
       epoch_id_(utils::GenerateUUID()),
       global_locker_(file_retainer_.AddLocker()) {
+  if (config_.durability.snapshot_wal_mode == Config::Durability::SnapshotWalMode::DISABLED &&
+      replication_role_ == ReplicationRole::MAIN) {
+    spdlog::warn(
+        "The instance has the MAIN replication role, but durability logs and snapshots are disabled. Please consider "
+        "enabling durability by using --storage-snapshot-interval-sec and --storage-wal-enabled flags because "
+        "without write-ahead logs this instance is not replicating any data.");
+  }
   if (config_.durability.snapshot_wal_mode != Config::Durability::SnapshotWalMode::DISABLED ||
       config_.durability.snapshot_on_exit || config_.durability.recover_on_startup) {
     // Create the directory initially to crash the database in case of
@@ -393,6 +416,16 @@ Storage::Storage(Config config)
   } else {
     commit_log_.emplace(timestamp_);
   }
+
+  if (config_.durability.restore_replicas_on_startup) {
+    spdlog::info("Replica's configuration will be stored and will be automatically restored in case of a crash.");
+    utils::EnsureDirOrDie(config_.durability.storage_directory / durability::kReplicationDirectory);
+    storage_ =
+        std::make_unique<kvstore::KVStore>(config_.durability.storage_directory / durability::kReplicationDirectory);
+    RestoreReplicas();
+  } else {
+    spdlog::warn("Replicas' configuration will NOT be stored. When the server restarts, replicas will be forgotten.");
+  }
 }
 
 Storage::~Storage() {
@@ -805,11 +838,11 @@ const std::string &Storage::Accessor::EdgeTypeToName(EdgeTypeId edge_type) const
   return storage_->EdgeTypeToName(edge_type);
 }
 
-LabelId Storage::Accessor::NameToLabel(const std::string_view &name) { return storage_->NameToLabel(name); }
+LabelId Storage::Accessor::NameToLabel(const std::string_view name) { return storage_->NameToLabel(name); }
 
-PropertyId Storage::Accessor::NameToProperty(const std::string_view &name) { return storage_->NameToProperty(name); }
+PropertyId Storage::Accessor::NameToProperty(const std::string_view name) { return storage_->NameToProperty(name); }
 
-EdgeTypeId Storage::Accessor::NameToEdgeType(const std::string_view &name) { return storage_->NameToEdgeType(name); }
+EdgeTypeId Storage::Accessor::NameToEdgeType(const std::string_view name) { return storage_->NameToEdgeType(name); }
 
 void Storage::Accessor::AdvanceCommand() { ++transaction_.command_id; }
 
@@ -1114,13 +1147,13 @@ const std::string &Storage::EdgeTypeToName(EdgeTypeId edge_type) const {
   return name_id_mapper_.IdToName(edge_type.AsUint());
 }
 
-LabelId Storage::NameToLabel(const std::string_view &name) { return LabelId::FromUint(name_id_mapper_.NameToId(name)); }
+LabelId Storage::NameToLabel(const std::string_view name) { return LabelId::FromUint(name_id_mapper_.NameToId(name)); }
 
-PropertyId Storage::NameToProperty(const std::string_view &name) {
+PropertyId Storage::NameToProperty(const std::string_view name) {
   return PropertyId::FromUint(name_id_mapper_.NameToId(name));
 }
 
-EdgeTypeId Storage::NameToEdgeType(const std::string_view &name) {
+EdgeTypeId Storage::NameToEdgeType(const std::string_view name) {
   return EdgeTypeId::FromUint(name_id_mapper_.NameToId(name));
 }
 
@@ -1875,40 +1908,77 @@ bool Storage::SetMainReplicationRole() {
 
 utils::BasicResult<Storage::RegisterReplicaError> Storage::RegisterReplica(
     std::string name, io::network::Endpoint endpoint, const replication::ReplicationMode replication_mode,
-    const replication::ReplicationClientConfig &config) {
+    const replication::RegistrationMode registration_mode, const replication::ReplicationClientConfig &config) {
   MG_ASSERT(replication_role_.load() == ReplicationRole::MAIN, "Only main instance can register a replica!");
 
   const bool name_exists = replication_clients_.WithLock([&](auto &clients) {
-    return std::any_of(clients.begin(), clients.end(), [&](auto &client) { return client->Name() == name; });
+    return std::any_of(clients.begin(), clients.end(), [&name](const auto &client) { return client->Name() == name; });
   });
 
   if (name_exists) {
     return RegisterReplicaError::NAME_EXISTS;
   }
 
-  MG_ASSERT(replication_mode == replication::ReplicationMode::SYNC || !config.timeout,
-            "Only SYNC mode can have a timeout set");
+  const auto end_point_exists = replication_clients_.WithLock([&endpoint](auto &clients) {
+    return std::any_of(clients.begin(), clients.end(),
+                       [&endpoint](const auto &client) { return client->Endpoint() == endpoint; });
+  });
+
+  if (end_point_exists) {
+    return RegisterReplicaError::END_POINT_EXISTS;
+  }
+
+  if (ShouldStoreAndRestoreReplicas()) {
+    auto data = replication::ReplicaStatusToJSON(
+        replication::ReplicaStatus{.name = name,
+                                   .ip_address = endpoint.address,
+                                   .port = endpoint.port,
+                                   .sync_mode = replication_mode,
+                                   .replica_check_frequency = config.replica_check_frequency,
+                                   .ssl = config.ssl});
+    if (!storage_->Put(name, data.dump())) {
+      spdlog::error("Error when saving replica {} in settings.", name);
+      return RegisterReplicaError::COULD_NOT_BE_PERSISTED;
+    }
+  }
 
   auto client = std::make_unique<ReplicationClient>(std::move(name), this, endpoint, replication_mode, config);
+
   if (client->State() == replication::ReplicaState::INVALID) {
-    return RegisterReplicaError::CONNECTION_FAILED;
+    if (replication::RegistrationMode::CAN_BE_INVALID != registration_mode) {
+      return RegisterReplicaError::CONNECTION_FAILED;
+    }
+
+    spdlog::warn("Connection failed when registering replica {}. Replica will still be registered.", client->Name());
   }
 
   return replication_clients_.WithLock([&](auto &clients) -> utils::BasicResult<Storage::RegisterReplicaError> {
     // Another thread could have added a client with same name while
     // we were connecting to this client.
     if (std::any_of(clients.begin(), clients.end(),
-                    [&](auto &other_client) { return client->Name() == other_client->Name(); })) {
+                    [&](const auto &other_client) { return client->Name() == other_client->Name(); })) {
       return RegisterReplicaError::NAME_EXISTS;
     }
 
+    if (std::any_of(clients.begin(), clients.end(),
+                    [&client](const auto &other_client) { return client->Endpoint() == other_client->Endpoint(); })) {
+      return RegisterReplicaError::END_POINT_EXISTS;
+    }
+
     clients.push_back(std::move(client));
     return {};
   });
 }
 
-bool Storage::UnregisterReplica(const std::string_view name) {
+bool Storage::UnregisterReplica(const std::string &name) {
   MG_ASSERT(replication_role_.load() == ReplicationRole::MAIN, "Only main instance can unregister a replica!");
+  if (ShouldStoreAndRestoreReplicas()) {
+    if (!storage_->Delete(name)) {
+      spdlog::error("Error when removing replica {} from settings.", name);
+      return false;
+    }
+  }
+
   return replication_clients_.WithLock([&](auto &clients) {
     return std::erase_if(clients, [&](const auto &client) { return client->Name() == name; });
   });
@@ -1931,10 +2001,10 @@ std::vector<Storage::ReplicaInfo> Storage::ReplicasInfo() {
   return replication_clients_.WithLock([](auto &clients) {
     std::vector<Storage::ReplicaInfo> replica_info;
     replica_info.reserve(clients.size());
-    std::transform(clients.begin(), clients.end(), std::back_inserter(replica_info),
-                   [](const auto &client) -> ReplicaInfo {
-                     return {client->Name(), client->Mode(), client->Timeout(), client->Endpoint(), client->State()};
-                   });
+    std::transform(
+        clients.begin(), clients.end(), std::back_inserter(replica_info), [](const auto &client) -> ReplicaInfo {
+          return {client->Name(), client->Mode(), client->Endpoint(), client->State(), client->GetTimestampInfo()};
+        });
     return replica_info;
   });
 }
@@ -1944,4 +2014,41 @@ void Storage::SetIsolationLevel(IsolationLevel isolation_level) {
   isolation_level_ = isolation_level;
 }
 
+void Storage::RestoreReplicas() {
+  MG_ASSERT(memgraph::storage::ReplicationRole::MAIN == GetReplicationRole());
+  if (!ShouldStoreAndRestoreReplicas()) {
+    return;
+  }
+  spdlog::info("Restoring replicas.");
+
+  for (const auto &[replica_name, replica_data] : *storage_) {
+    spdlog::info("Restoring replica {}.", replica_name);
+
+    const auto maybe_replica_status = replication::JSONToReplicaStatus(nlohmann::json::parse(replica_data));
+    if (!maybe_replica_status.has_value()) {
+      LOG_FATAL("Cannot parse previously saved configuration of replica {}.", replica_name);
+    }
+
+    auto replica_status = *maybe_replica_status;
+    MG_ASSERT(replica_status.name == replica_name, "Expected replica name is '{}', but got '{}'", replica_status.name,
+              replica_name);
+
+    auto ret =
+        RegisterReplica(std::move(replica_status.name), {std::move(replica_status.ip_address), replica_status.port},
+                        replica_status.sync_mode, replication::RegistrationMode::CAN_BE_INVALID,
+                        {
+                            .replica_check_frequency = replica_status.replica_check_frequency,
+                            .ssl = replica_status.ssl,
+                        });
+
+    if (ret.HasError()) {
+      MG_ASSERT(RegisterReplicaError::CONNECTION_FAILED != ret.GetError());
+      LOG_FATAL("Failure when restoring replica {}: {}.", replica_name, RegisterReplicaErrorToString(ret.GetError()));
+    }
+    spdlog::info("Replica {} restored.", replica_name);
+  }
+}
+
+bool Storage::ShouldStoreAndRestoreReplicas() const { return nullptr != storage_; }
+
 }  // namespace memgraph::storage

src/storage/v2/storage.hpp

@@ -18,6 +18,7 @@
 #include <variant>
 
 #include "io/network/endpoint.hpp"
+#include "kvstore/kvstore.hpp"
 #include "storage/v2/commit_log.hpp"
 #include "storage/v2/config.hpp"
 #include "storage/v2/constraints.hpp"
@@ -283,13 +284,13 @@ class Storage final {
     const std::string &EdgeTypeToName(EdgeTypeId edge_type) const;
 
     /// @throw std::bad_alloc if unable to insert a new mapping
-    LabelId NameToLabel(const std::string_view &name);
+    LabelId NameToLabel(std::string_view name);
 
     /// @throw std::bad_alloc if unable to insert a new mapping
-    PropertyId NameToProperty(const std::string_view &name);
+    PropertyId NameToProperty(std::string_view name);
 
     /// @throw std::bad_alloc if unable to insert a new mapping
-    EdgeTypeId NameToEdgeType(const std::string_view &name);
+    EdgeTypeId NameToEdgeType(std::string_view name);
 
     bool LabelIndexExists(LabelId label) const { return storage_->indices_.label_index.IndexExists(label); }
 
@@ -343,13 +344,13 @@ class Storage final {
   const std::string &EdgeTypeToName(EdgeTypeId edge_type) const;
 
   /// @throw std::bad_alloc if unable to insert a new mapping
-  LabelId NameToLabel(const std::string_view &name);
+  LabelId NameToLabel(std::string_view name);
 
   /// @throw std::bad_alloc if unable to insert a new mapping
-  PropertyId NameToProperty(const std::string_view &name);
+  PropertyId NameToProperty(std::string_view name);
 
   /// @throw std::bad_alloc if unable to insert a new mapping
-  EdgeTypeId NameToEdgeType(const std::string_view &name);
+  EdgeTypeId NameToEdgeType(std::string_view name);
 
   /// @throw std::bad_alloc
   bool CreateIndex(LabelId label, std::optional<uint64_t> desired_commit_timestamp = {});
@@ -411,26 +412,36 @@ class Storage final {
 
   bool SetMainReplicationRole();
 
-  enum class RegisterReplicaError : uint8_t { NAME_EXISTS, CONNECTION_FAILED };
+  enum class RegisterReplicaError : uint8_t {
+    NAME_EXISTS,
+    END_POINT_EXISTS,
+    CONNECTION_FAILED,
+    COULD_NOT_BE_PERSISTED
+  };
 
   /// @pre The instance should have a MAIN role
   /// @pre Timeout can only be set for SYNC replication
   utils::BasicResult<RegisterReplicaError, void> RegisterReplica(
       std::string name, io::network::Endpoint endpoint, replication::ReplicationMode replication_mode,
-      const replication::ReplicationClientConfig &config = {});
+      replication::RegistrationMode registration_mode, const replication::ReplicationClientConfig &config = {});
   /// @pre The instance should have a MAIN role
-  bool UnregisterReplica(std::string_view name);
+  bool UnregisterReplica(const std::string &name);
 
   std::optional<replication::ReplicaState> GetReplicaState(std::string_view name);
 
   ReplicationRole GetReplicationRole() const;
 
+  struct TimestampInfo {
+    uint64_t current_timestamp_of_replica;
+    uint64_t current_number_of_timestamp_behind_master;
+  };
+
   struct ReplicaInfo {
     std::string name;
     replication::ReplicationMode mode;
-    std::optional<double> timeout;
     io::network::Endpoint endpoint;
     replication::ReplicaState state;
+    TimestampInfo timestamp_info;
   };
 
   std::vector<ReplicaInfo> ReplicasInfo();
@@ -469,6 +480,10 @@ class Storage final {
 
   uint64_t CommitTimestamp(std::optional<uint64_t> desired_commit_timestamp = {});
 
+  void RestoreReplicas();
+
+  bool ShouldStoreAndRestoreReplicas() const;
+
   // Main storage lock.
   //
   // Accessors take a shared lock when starting, so it is possible to block
@@ -529,6 +544,7 @@ class Storage final {
   std::filesystem::path wal_directory_;
   std::filesystem::path lock_file_path_;
   utils::OutputFile lock_file_handle_;
+  std::unique_ptr<kvstore::KVStore> storage_;
 
   utils::Scheduler snapshot_runner_;
   utils::SpinLock snapshot_lock_;

src/utils/file.cpp

@@ -370,7 +370,7 @@ void OutputFile::Write(const uint8_t *data, size_t size) {
 }
 
 void OutputFile::Write(const char *data, size_t size) { Write(reinterpret_cast<const uint8_t *>(data), size); }
-void OutputFile::Write(const std::string_view &data) { Write(data.data(), data.size()); }
+void OutputFile::Write(const std::string_view data) { Write(data.data(), data.size()); }
 
 size_t OutputFile::SeekFile(const Position position, const ssize_t offset) {
   int whence;

src/utils/file.hpp

@@ -209,7 +209,7 @@ class OutputFile {
   /// the program.
   void Write(const uint8_t *data, size_t size);
   void Write(const char *data, size_t size);
-  void Write(const std::string_view &data);
+  void Write(std::string_view data);
 
   /// This method gets the current absolute position in the file. On failure and
   /// misuse it crashes the program.

src/utils/fnv.hpp

@@ -16,7 +16,7 @@
 
 namespace memgraph::utils {
 
-inline uint64_t Fnv(const std::string_view &s) {
+inline uint64_t Fnv(const std::string_view s) {
   // fnv1a is recommended so use it as the default implementation.
   uint64_t hash = 14695981039346656037UL;
 

src/utils/string.hpp

@@ -29,7 +29,7 @@
 namespace memgraph::utils {
 
 /** Remove whitespace characters from the start of a string. */
-inline std::string_view LTrim(const std::string_view &s) {
+inline std::string_view LTrim(const std::string_view s) {
   size_t start = 0;
   while (start < s.size() && isspace(s[start])) {
     ++start;
@@ -38,7 +38,7 @@ inline std::string_view LTrim(const std::string_view &s) {
 }
 
 /** Remove characters found in `chars` from the start of a string. */
-inline std::string_view LTrim(const std::string_view &s, const std::string_view &chars) {
+inline std::string_view LTrim(const std::string_view s, const std::string_view chars) {
   size_t start = 0;
   while (start < s.size() && chars.find(s[start]) != std::string::npos) {
     ++start;
@@ -47,7 +47,7 @@ inline std::string_view LTrim(const std::string_view &s, const std::string_view
 }
 
 /** Remove whitespace characters from the end of a string. */
-inline std::string_view RTrim(const std::string_view &s) {
+inline std::string_view RTrim(const std::string_view s) {
   size_t count = s.size();
   while (count > static_cast<size_t>(0) && isspace(s[count - 1])) {
     --count;
@@ -56,7 +56,7 @@ inline std::string_view RTrim(const std::string_view &s) {
 }
 
 /** Remove characters found in `chars` from the end of a string. */
-inline std::string_view RTrim(const std::string_view &s, const std::string_view &chars) {
+inline std::string_view RTrim(const std::string_view s, const std::string_view chars) {
   size_t count = s.size();
   while (count > static_cast<size_t>(0) && chars.find(s[count - 1]) != std::string::npos) {
     --count;
@@ -65,7 +65,7 @@ inline std::string_view RTrim(const std::string_view &s, const std::string_view
 }
 
 /** Remove whitespace characters from the start and from the end of a string. */
-inline std::string_view Trim(const std::string_view &s) {
+inline std::string_view Trim(const std::string_view s) {
   size_t start = 0;
   size_t count = s.size();
   while (start < s.size() && isspace(s[start])) {
@@ -78,7 +78,7 @@ inline std::string_view Trim(const std::string_view &s) {
 }
 
 /** Remove characters found in `chars` from the start and the end of `s`. */
-inline std::string_view Trim(const std::string_view &s, const std::string_view &chars) {
+inline std::string_view Trim(const std::string_view s, const std::string_view chars) {
   size_t start = 0;
   size_t count = s.size();
   while (start < s.size() && chars.find(s[start]) != std::string::npos) {
@@ -97,7 +97,7 @@ inline std::string_view Trim(const std::string_view &s, const std::string_view &
  */
 template <class TAllocator>
 std::basic_string<char, std::char_traits<char>, TAllocator> *ToLowerCase(
-    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view &s) {
+    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view s) {
   out->resize(s.size());
   std::transform(s.begin(), s.end(), out->begin(), [](char c) { return tolower(c); });
   return out;
@@ -107,7 +107,7 @@ std::basic_string<char, std::char_traits<char>, TAllocator> *ToLowerCase(
  * Lowercase all characters of a string.
  * Transformation is locale independent.
  */
-inline std::string ToLowerCase(const std::string_view &s) {
+inline std::string ToLowerCase(const std::string_view s) {
   std::string res;
   ToLowerCase(&res, s);
   return res;
@@ -120,7 +120,7 @@ inline std::string ToLowerCase(const std::string_view &s) {
  */
 template <class TAllocator>
 std::basic_string<char, std::char_traits<char>, TAllocator> *ToUpperCase(
-    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view &s) {
+    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view s) {
   out->resize(s.size());
   std::transform(s.begin(), s.end(), out->begin(), [](char c) { return toupper(c); });
   return out;
@@ -130,7 +130,7 @@ std::basic_string<char, std::char_traits<char>, TAllocator> *ToUpperCase(
  * Uppercase all characters of a string and store the result in `out`.
  * Transformation is locale independent.
  */
-inline std::string ToUpperCase(const std::string_view &s) {
+inline std::string ToUpperCase(const std::string_view s) {
   std::string res;
   ToUpperCase(&res, s);
   return res;
@@ -143,7 +143,7 @@ inline std::string ToUpperCase(const std::string_view &s) {
 template <class TCollection, class TAllocator>
 std::basic_string<char, std::char_traits<char>, TAllocator> *Join(
     std::basic_string<char, std::char_traits<char>, TAllocator> *out, const TCollection &strings,
-    const std::string_view &separator) {
+    const std::string_view separator) {
   out->clear();
   if (strings.empty()) return out;
   int64_t total_size = 0;
@@ -163,7 +163,7 @@ std::basic_string<char, std::char_traits<char>, TAllocator> *Join(
 /**
  * Join the `strings` collection separated by a given separator.
  */
-inline std::string Join(const std::vector<std::string> &strings, const std::string_view &separator) {
+inline std::string Join(const std::vector<std::string> &strings, const std::string_view separator) {
   std::string res;
   Join(&res, strings, separator);
   return res;
@@ -175,8 +175,8 @@ inline std::string Join(const std::vector<std::string> &strings, const std::stri
  */
 template <class TAllocator>
 std::basic_string<char, std::char_traits<char>, TAllocator> *Replace(
-    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view &src,
-    const std::string_view &match, const std::string_view &replacement) {
+    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view src,
+    const std::string_view match, const std::string_view replacement) {
   // TODO: This could be implemented much more efficiently.
   *out = src;
   for (size_t pos = out->find(match); pos != std::string::npos; pos = out->find(match, pos + replacement.size())) {
@@ -186,8 +186,8 @@ std::basic_string<char, std::char_traits<char>, TAllocator> *Replace(
 }
 
 /** Replace all occurrences of `match` in `src` with `replacement`. */
-inline std::string Replace(const std::string_view &src, const std::string_view &match,
-                           const std::string_view &replacement) {
+inline std::string Replace(const std::string_view src, const std::string_view match,
+                           const std::string_view replacement) {
   std::string res;
   Replace(&res, src, match, replacement);
   return res;
@@ -200,8 +200,8 @@ inline std::string Replace(const std::string_view &src, const std::string_view &
  * @return pointer to `out`.
  */
 template <class TString, class TAllocator>
-std::vector<TString, TAllocator> *Split(std::vector<TString, TAllocator> *out, const std::string_view &src,
-                                        const std::string_view &delimiter, int splits = -1) {
+std::vector<TString, TAllocator> *Split(std::vector<TString, TAllocator> *out, const std::string_view src,
+                                        const std::string_view delimiter, int splits = -1) {
   out->clear();
   if (src.empty()) return out;
   size_t index = 0;
@@ -220,7 +220,7 @@ std::vector<TString, TAllocator> *Split(std::vector<TString, TAllocator> *out, c
  * The vector will have at most `splits` + 1 elements. Negative value of
  * `splits` indicates to perform all possible splits.
  */
-inline std::vector<std::string> Split(const std::string_view &src, const std::string_view &delimiter, int splits = -1) {
+inline std::vector<std::string> Split(const std::string_view src, const std::string_view delimiter, int splits = -1) {
   std::vector<std::string> res;
   Split(&res, src, delimiter, splits);
   return res;
@@ -234,7 +234,7 @@ inline std::vector<std::string> Split(const std::string_view &src, const std::st
  * @return pointer to `out`.
  */
 template <class TString, class TAllocator>
-std::vector<TString, TAllocator> *Split(std::vector<TString, TAllocator> *out, const std::string_view &src) {
+std::vector<TString, TAllocator> *Split(std::vector<TString, TAllocator> *out, const std::string_view src) {
   out->clear();
   if (src.empty()) return out;
   // TODO: Investigate how much regex allocate and perhaps replace with custom
@@ -256,7 +256,7 @@ std::vector<TString, TAllocator> *Split(std::vector<TString, TAllocator> *out, c
  * Additionally, the result will not contain empty strings at the start or end
  * as if the string was trimmed before splitting.
  */
-inline std::vector<std::string> Split(const std::string_view &src) {
+inline std::vector<std::string> Split(const std::string_view src) {
   std::vector<std::string> res;
   Split(&res, src);
   return res;
@@ -271,8 +271,8 @@ inline std::vector<std::string> Split(const std::string_view &src) {
  * @return pointer to `out`.
  */
 template <class TString, class TAllocator>
-std::vector<TString, TAllocator> *RSplit(std::vector<TString, TAllocator> *out, const std::string_view &src,
-                                         const std::string_view &delimiter, int splits = -1) {
+std::vector<TString, TAllocator> *RSplit(std::vector<TString, TAllocator> *out, const std::string_view src,
+                                         const std::string_view delimiter, int splits = -1) {
   out->clear();
   if (src.empty()) return out;
   size_t index = src.size();
@@ -295,8 +295,7 @@ std::vector<TString, TAllocator> *RSplit(std::vector<TString, TAllocator> *out,
  * have at most `splits` + 1 elements. Negative value of `splits` indicates to
  * perform all possible splits.
  */
-inline std::vector<std::string> RSplit(const std::string_view &src, const std::string_view &delimiter,
-                                       int splits = -1) {
+inline std::vector<std::string> RSplit(const std::string_view src, const std::string_view delimiter, int splits = -1) {
   std::vector<std::string> res;
   RSplit(&res, src, delimiter, splits);
   return res;
@@ -309,7 +308,7 @@ inline std::vector<std::string> RSplit(const std::string_view &src, const std::s
  *
  * @throw BasicException if unable to parse the whole string.
  */
-inline int64_t ParseInt(const std::string_view &s) {
+inline int64_t ParseInt(const std::string_view s) {
   // stol would be nicer but it uses current locale so we shouldn't use it.
   int64_t t = 0;
   // NOTE: Constructing std::istringstream will make a copy of the string, which
@@ -336,7 +335,7 @@ inline int64_t ParseInt(const std::string_view &s) {
  *
  * @throw BasicException if unable to parse the whole string.
  */
-inline double ParseDouble(const std::string_view &s) {
+inline double ParseDouble(const std::string_view s) {
   // stod would be nicer but it uses current locale so we shouldn't use it.
   double t = 0.0;
   // NOTE: Constructing std::istringstream will make a copy of the string, which
@@ -357,17 +356,17 @@ inline double ParseDouble(const std::string_view &s) {
 }
 
 /** Check if the given string `s` ends with the given `suffix`. */
-inline bool EndsWith(const std::string_view &s, const std::string_view &suffix) {
+inline bool EndsWith(const std::string_view s, const std::string_view suffix) {
   return s.size() >= suffix.size() && s.compare(s.size() - suffix.size(), std::string::npos, suffix) == 0;
 }
 
 /** Check if the given string `s` starts with the given `prefix`. */
-inline bool StartsWith(const std::string_view &s, const std::string_view &prefix) {
+inline bool StartsWith(const std::string_view s, const std::string_view prefix) {
   return s.size() >= prefix.size() && s.compare(0, prefix.size(), prefix) == 0;
 }
 
 /** Perform case-insensitive string equality test. */
-inline bool IEquals(const std::string_view &lhs, const std::string_view &rhs) {
+inline bool IEquals(const std::string_view lhs, const std::string_view rhs) {
   if (lhs.size() != rhs.size()) return false;
   for (size_t i = 0; i < lhs.size(); ++i) {
     if (tolower(lhs[i]) != tolower(rhs[i])) return false;
@@ -406,7 +405,7 @@ inline std::string RandomString(size_t length) {
  */
 template <class TAllocator>
 std::basic_string<char, std::char_traits<char>, TAllocator> *Escape(
-    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view &src) {
+    std::basic_string<char, std::char_traits<char>, TAllocator> *out, const std::string_view src) {
   out->clear();
   out->reserve(src.size() + 2);
   out->append(1, '"');
@@ -433,7 +432,7 @@ std::basic_string<char, std::char_traits<char>, TAllocator> *Escape(
 }
 
 /** Escape all whitespace and quotation characters in the given string. */
-inline std::string Escape(const std::string_view &src) {
+inline std::string Escape(const std::string_view src) {
   std::string res;
   Escape(&res, src);
   return res;
@@ -445,7 +444,7 @@ inline std::string Escape(const std::string_view &src) {
  * clamped to a valid interval. Therefore, this function never throws
  * std::out_of_range, unlike std::basic_string::substr.
  */
-inline std::string_view Substr(const std::string_view &string, size_t pos = 0, size_t count = std::string::npos) {
+inline std::string_view Substr(const std::string_view string, size_t pos = 0, size_t count = std::string::npos) {
   if (pos >= string.size()) return std::string_view(string.data(), 0);
   auto len = std::min(string.size() - pos, count);
   return string.substr(pos, len);

src/utils/temporal.cpp

@@ -15,6 +15,7 @@
 #include <chrono>
 #include <ctime>
 #include <limits>
+#include <string>
 #include <string_view>
 
 #include "utils/exceptions.hpp"
@@ -175,6 +176,10 @@ int64_t Date::MicrosecondsSinceEpoch() const {
 
 int64_t Date::DaysSinceEpoch() const { return utils::DaysSinceEpoch(year, month, day).count(); }
 
+std::string Date::ToString() const {
+  return fmt::format("{:0>4}-{:0>2}-{:0>2}", year, static_cast<int>(month), static_cast<int>(day));
+}
+
 size_t DateHash::operator()(const Date &date) const {
   utils::HashCombine<uint64_t, uint64_t> hasher;
   size_t result = hasher(0, date.year);
@@ -377,6 +382,15 @@ int64_t LocalTime::NanosecondsSinceEpoch() const {
   return chrono::duration_cast<chrono::nanoseconds>(SumLocalTimeParts()).count();
 }
 
+std::string LocalTime::ToString() const {
+  using milli = std::chrono::milliseconds;
+  using micro = std::chrono::microseconds;
+  const auto subseconds = milli(millisecond) + micro(microsecond);
+
+  return fmt::format("{:0>2}:{:0>2}:{:0>2}.{:0>6}", static_cast<int>(hour), static_cast<int>(minute),
+                     static_cast<int>(second), subseconds.count());
+}
+
 size_t LocalTimeHash::operator()(const LocalTime &local_time) const {
   utils::HashCombine<uint64_t, uint64_t> hasher;
   size_t result = hasher(0, local_time.hour);
@@ -486,6 +500,8 @@ int64_t LocalDateTime::SubSecondsAsNanoseconds() const {
   return (milli_as_nanos + micros_as_nanos).count();
 }
 
+std::string LocalDateTime::ToString() const { return date.ToString() + 'T' + local_time.ToString(); }
+
 LocalDateTime::LocalDateTime(const DateParameters &date_parameters, const LocalTimeParameters &local_time_parameters)
     : date(date_parameters), local_time(local_time_parameters) {}
 
@@ -699,6 +715,23 @@ int64_t Duration::SubSecondsAsNanoseconds() const {
   return chrono::duration_cast<chrono::nanoseconds>(micros - secs).count();
 }
 
+std::string Duration::ToString() const {
+  // Format [nD]T[nH]:[nM]:[nS].
+  namespace chrono = std::chrono;
+  auto micros = chrono::microseconds(microseconds);
+  const auto dd = GetAndSubtractDuration<chrono::days>(micros);
+  const auto h = GetAndSubtractDuration<chrono::hours>(micros);
+  const auto m = GetAndSubtractDuration<chrono::minutes>(micros);
+  const auto s = GetAndSubtractDuration<chrono::seconds>(micros);
+
+  auto first_half = fmt::format("P{}DT{}H{}M", dd, h, m);
+  auto second_half = fmt::format("{}.{:0>6}S", s, std::abs(micros.count()));
+  if (s == 0 && micros.count() < 0) {
+    return first_half + '-' + second_half;
+  }
+  return first_half + second_half;
+}
+
 Duration Duration::operator-() const {
   if (microseconds == std::numeric_limits<decltype(microseconds)>::min()) [[unlikely]] {
     throw temporal::InvalidArgumentException("Duration arithmetic overflows");

src/utils/temporal.hpp

@@ -87,20 +87,9 @@ struct Duration {
   int64_t SubDaysAsNanoseconds() const;
   int64_t SubSecondsAsNanoseconds() const;
 
-  friend std::ostream &operator<<(std::ostream &os, const Duration &dur) {
-    // Format [nD]T[nH]:[nM]:[nS].
-    namespace chrono = std::chrono;
-    auto micros = chrono::microseconds(dur.microseconds);
-    const auto dd = GetAndSubtractDuration<chrono::days>(micros);
-    const auto h = GetAndSubtractDuration<chrono::hours>(micros);
-    const auto m = GetAndSubtractDuration<chrono::minutes>(micros);
-    const auto s = GetAndSubtractDuration<chrono::seconds>(micros);
-    os << fmt::format("P{}DT{}H{}M", dd, h, m);
-    if (s == 0 && micros.count() < 0) {
-      os << '-';
-    }
-    return os << fmt::format("{}.{:0>6}S", s, std::abs(micros.count()));
-  }
+  std::string ToString() const;
+
+  friend std::ostream &operator<<(std::ostream &os, const Duration &dur) { return os << dur.ToString(); }
 
   Duration operator-() const;
 
@@ -155,13 +144,11 @@ struct Date {
   explicit Date(int64_t microseconds);
   explicit Date(const DateParameters &date_parameters);
 
-  friend std::ostream &operator<<(std::ostream &os, const Date &date) {
-    return os << fmt::format("{:0>2}-{:0>2}-{:0>2}", date.year, static_cast<int>(date.month),
-                             static_cast<int>(date.day));
-  }
+  friend std::ostream &operator<<(std::ostream &os, const Date &date) { return os << date.ToString(); }
 
   int64_t MicrosecondsSinceEpoch() const;
   int64_t DaysSinceEpoch() const;
+  std::string ToString() const;
 
   friend Date operator+(const Date &date, const Duration &dur) {
     namespace chrono = std::chrono;
@@ -217,17 +204,11 @@ struct LocalTime {
   // Epoch means the start of the day, i,e, midnight
   int64_t MicrosecondsSinceEpoch() const;
   int64_t NanosecondsSinceEpoch() const;
+  std::string ToString() const;
 
   auto operator<=>(const LocalTime &) const = default;
 
-  friend std::ostream &operator<<(std::ostream &os, const LocalTime &lt) {
-    namespace chrono = std::chrono;
-    using milli = chrono::milliseconds;
-    using micro = chrono::microseconds;
-    const auto subseconds = milli(lt.millisecond) + micro(lt.microsecond);
-    return os << fmt::format("{:0>2}:{:0>2}:{:0>2}.{:0>6}", static_cast<int>(lt.hour), static_cast<int>(lt.minute),
-                             static_cast<int>(lt.second), subseconds.count());
-  }
+  friend std::ostream &operator<<(std::ostream &os, const LocalTime &lt) { return os << lt.ToString(); }
 
   friend LocalTime operator+(const LocalTime &local_time, const Duration &dur) {
     namespace chrono = std::chrono;
@@ -279,13 +260,11 @@ struct LocalDateTime {
   int64_t MicrosecondsSinceEpoch() const;
   int64_t SecondsSinceEpoch() const;  // seconds since epoch
   int64_t SubSecondsAsNanoseconds() const;
+  std::string ToString() const;
 
   auto operator<=>(const LocalDateTime &) const = default;
 
-  friend std::ostream &operator<<(std::ostream &os, const LocalDateTime &ldt) {
-    os << ldt.date << 'T' << ldt.local_time;
-    return os;
-  }
+  friend std::ostream &operator<<(std::ostream &os, const LocalDateTime &ldt) { return os << ldt.ToString(); }
 
   friend LocalDateTime operator+(const LocalDateTime &dt, const Duration &dur) {
     const auto local_date_time_as_duration = Duration(dt.MicrosecondsSinceEpoch());