9 Commits

Author SHA1 Message Date
Chris Mumford
031ae8e2db Test cmd markdown 2019-04-12 12:02:21 -07:00
Chris Mumford
5a2a472741 Fixed missing std namespaces and make_unique.
cout/endl were missing the std namespace. Also std::make_unique
was used inadvertently which is part of C++14 and only C++11
is currently supported.

PiperOrigin-RevId: 243221310
2019-04-12 01:11:25 -07:00
Chris Mumford
08e771901f Simplify issue320_test.
Use std::unique_ptr to simplify issue320_test.

PiperOrigin-RevId: 243190799
2019-04-12 00:17:13 -07:00
Chris Mumford
65e86f75ea Fix formatting of recent snapshot compaction fix.
Fix variable names, line lengths, namespace use, and a few other
minor issues to conform to Google C++ style guide.

PiperOrigin-RevId: 243187729
2019-04-12 00:17:03 -07:00
Chris Mumford
7711e76766 Merge pull request #339 from richcole-at-amazon:master
PiperOrigin-RevId: 243156105
2019-04-12 00:16:52 -07:00
Chris Mumford
71ed7c401e Fixed typo in comment in version_set.h.
Flagged by presubmit check.

PiperOrigin-RevId: 243118632
2019-04-12 00:16:41 -07:00
Victor Costan
09fa8868db Align version/soversion CMake setup closer with other repositories.
PiperOrigin-RevId: 241432456
2019-04-01 17:37:45 -07:00
Richard Cole
20fb601aa9 Fix snapshot compaction bug
Closes google/leveldb#320

During compaction it was possible that records from a block b1=(l1,u1)
would be pushed down from level i to level i+1. If there is a block
b2=(l2,u2) at level i with k1 = user_key(u1) = user_key(l2) then
a subsequent search for k1 will yield the record l2 which has a smaller
sequence number than u1 because the sort order for records sorts
increasing by user key but decreaing by sequence number.

This change add a call to a new function AddBoundaryInputs to
SetupOtherInputs. AddBoundaryInputs searches for a block b2 matching the
criteria above and adds it to the set of files to be compacted. Whenever
AddBoundaryInputs is called it is important that the compaction fileset
in level i+1 (known as c->inputs_[1] in the code) be recomputed. Each
call to AddBoundaryInputs is followed by a call to GetOverlappingInputs.

SetupOtherInputs is called on both manual and automated compaction
passes. It is called for both level zero and for levels greater than 0.

The original change posted in https://github.com/google/leveldb/pull/339
has been modified to also include changed made by Chris Mumford<cmumford@google.com>
in 4b72cb14f8

  1. Releasing snapshots during test cleanup to avoid
     memory leak warnings.
  2. Refactored test to use testutil.h to be in line
     with other issue tests and to create the test
     database in the correct temporary location.
  3. Added copyright banner.

  Otherwise, just minor formatting and limiting character
  width to 80 characters.

Additionally the change was rebased on top of current master and
changes previously made to the Makefile were ported to the
CMakeLists.txt.

Testing Done:

  A test program (issue320_test) was constructed that performs mutations
  while snapshots are active. issue320_test fails without this bug fix
  after 64k writes. It passes with this bug fix. It was run with 200M
  writes and passed.

  Unit tests were written for the new function that was added to the
  code. Make test was run and seen to pass.

Signed-off-by: Richard Cole <richcole@amazon.com>
2019-04-01 13:10:09 -07:00
leveldb Team
37300aa54b Restore soname versioning with CMake build
Before:

$ readelf -d build/libleveldb.so | grep soname
 0x000000000000000e (SONAME)             Library soname: [libleveldb.so]

After:
$ readelf -d build/libleveldb.so | grep soname
 0x000000000000000e (SONAME)             Library soname: [libleveldb.so.1]

This matches the soname from v1.20.

PiperOrigin-RevId: 241334113
2019-04-01 09:04:40 -07:00
6 changed files with 375 additions and 8 deletions

View File

@@ -213,6 +213,10 @@ target_include_directories(leveldb
$<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/include>
$<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>
)
set_target_properties(leveldb
PROPERTIES VERSION ${PROJECT_VERSION} SOVERSION ${PROJECT_VERSION_MAJOR})
target_compile_definitions(leveldb
PRIVATE
# Used by include/export.h when building shared libraries.
@@ -297,6 +301,7 @@ if(LEVELDB_BUILD_TESTS)
leveldb_test("${PROJECT_SOURCE_DIR}/issues/issue178_test.cc")
leveldb_test("${PROJECT_SOURCE_DIR}/issues/issue200_test.cc")
leveldb_test("${PROJECT_SOURCE_DIR}/issues/issue320_test.cc")
leveldb_test("${PROJECT_SOURCE_DIR}/util/env_test.cc")
leveldb_test("${PROJECT_SOURCE_DIR}/util/status_test.cc")

View File

@@ -44,20 +44,20 @@ cmake -DCMAKE_BUILD_TYPE=Release .. && cmake --build .
First generate the Visual Studio 2017 project/solution files:
```bash
mkdir -p build
```cmd
mkdir build
cd build
cmake -G "Visual Studio 15" ..
```
The default default will build for x86. For 64-bit run:
```bash
```cmd
cmake -G "Visual Studio 15 Win64" ..
```
To compile the Windows solution from the command-line:
```bash
```cmd
devenv /build Debug leveldb.sln
```

View File

@@ -1347,9 +1347,90 @@ Compaction* VersionSet::PickCompaction() {
return c;
}
// Finds the largest key in a vector of files. Returns true if files it not
// empty.
bool FindLargestKey(const InternalKeyComparator& icmp,
const std::vector<FileMetaData*>& files,
InternalKey* largest_key) {
if (files.empty()) {
return false;
}
*largest_key = files[0]->largest;
for (size_t i = 1; i < files.size(); ++i) {
FileMetaData* f = files[i];
if (icmp.Compare(f->largest, *largest_key) > 0) {
*largest_key = f->largest;
}
}
return true;
}
// Finds minimum file b2=(l2, u2) in level file for which l2 > u1 and
// user_key(l2) = user_key(u1)
FileMetaData* FindSmallestBoundaryFile(
const InternalKeyComparator& icmp,
const std::vector<FileMetaData*>& level_files,
const InternalKey& largest_key) {
const Comparator* user_cmp = icmp.user_comparator();
FileMetaData* smallest_boundary_file = nullptr;
for (size_t i = 0; i < level_files.size(); ++i) {
FileMetaData* f = level_files[i];
if (icmp.Compare(f->smallest, largest_key) > 0 &&
user_cmp->Compare(f->smallest.user_key(), largest_key.user_key()) ==
0) {
if (smallest_boundary_file == nullptr ||
icmp.Compare(f->smallest, smallest_boundary_file->smallest) < 0) {
smallest_boundary_file = f;
}
}
}
return smallest_boundary_file;
}
// Extracts the largest file b1 from |compaction_files| and then searches for a
// b2 in |level_files| for which user_key(u1) = user_key(l2). If it finds such a
// file b2 (known as a boundary file) it adds it to |compaction_files| and then
// searches again using this new upper bound.
//
// If there are two blocks, b1=(l1, u1) and b2=(l2, u2) and
// user_key(u1) = user_key(l2), and if we compact b1 but not b2 then a
// subsequent get operation will yield an incorrect result because it will
// return the record from b2 in level i rather than from b1 because it searches
// level by level for records matching the supplied user key.
//
// parameters:
// in level_files: List of files to search for boundary files.
// in/out compaction_files: List of files to extend by adding boundary files.
void AddBoundaryInputs(const InternalKeyComparator& icmp,
const std::vector<FileMetaData*>& level_files,
std::vector<FileMetaData*>* compaction_files) {
InternalKey largest_key;
// Quick return if compaction_files is empty.
if (!FindLargestKey(icmp, *compaction_files, &largest_key)) {
return;
}
bool continue_searching = true;
while (continue_searching) {
FileMetaData* smallest_boundary_file =
FindSmallestBoundaryFile(icmp, level_files, largest_key);
// If a boundary file was found advance largest_key, otherwise we're done.
if (smallest_boundary_file != NULL) {
compaction_files->push_back(smallest_boundary_file);
largest_key = smallest_boundary_file->largest;
} else {
continue_searching = false;
}
}
}
void VersionSet::SetupOtherInputs(Compaction* c) {
const int level = c->level();
InternalKey smallest, largest;
AddBoundaryInputs(icmp_, current_->files_[level], &c->inputs_[0]);
GetRange(c->inputs_[0], &smallest, &largest);
current_->GetOverlappingInputs(level+1, &smallest, &largest, &c->inputs_[1]);
@@ -1363,6 +1444,7 @@ void VersionSet::SetupOtherInputs(Compaction* c) {
if (!c->inputs_[1].empty()) {
std::vector<FileMetaData*> expanded0;
current_->GetOverlappingInputs(level, &all_start, &all_limit, &expanded0);
AddBoundaryInputs(icmp_, current_->files_[level], &expanded0);
const int64_t inputs0_size = TotalFileSize(c->inputs_[0]);
const int64_t inputs1_size = TotalFileSize(c->inputs_[1]);
const int64_t expanded0_size = TotalFileSize(expanded0);

View File

@@ -375,7 +375,7 @@ class Compaction {
// Each compaction reads inputs from "level_" and "level_+1"
std::vector<FileMetaData*> inputs_[2]; // The two sets of inputs
// State used to check for number of of overlapping grandparent files
// State used to check for number of overlapping grandparent files
// (parent == level_ + 1, grandparent == level_ + 2)
std::vector<FileMetaData*> grandparents_;
size_t grandparent_index_; // Index in grandparent_starts_

View File

@@ -172,8 +172,160 @@ TEST(FindFileTest, OverlappingFiles) {
ASSERT_TRUE(Overlaps("600", "700"));
}
void AddBoundaryInputs(const InternalKeyComparator& icmp,
const std::vector<FileMetaData*>& level_files,
std::vector<FileMetaData*>* compaction_files);
class AddBoundaryInputsTest {
public:
std::vector<FileMetaData*> level_files_;
std::vector<FileMetaData*> compaction_files_;
std::vector<FileMetaData*> all_files_;
InternalKeyComparator icmp_;
AddBoundaryInputsTest() : icmp_(BytewiseComparator()){};
~AddBoundaryInputsTest() {
for (size_t i = 0; i < all_files_.size(); ++i) {
delete all_files_[i];
}
all_files_.clear();
};
FileMetaData* CreateFileMetaData(uint64_t number, InternalKey smallest,
InternalKey largest) {
FileMetaData* f = new FileMetaData();
f->number = number;
f->smallest = smallest;
f->largest = largest;
all_files_.push_back(f);
return f;
}
};
TEST(AddBoundaryInputsTest, TestEmptyFileSets) {
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_TRUE(compaction_files_.empty());
ASSERT_TRUE(level_files_.empty());
}
TEST(AddBoundaryInputsTest, TestEmptyLevelFiles) {
FileMetaData* f1 =
CreateFileMetaData(1, InternalKey("100", 2, kTypeValue),
InternalKey(InternalKey("100", 1, kTypeValue)));
compaction_files_.push_back(f1);
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_EQ(1, compaction_files_.size());
ASSERT_EQ(f1, compaction_files_[0]);
ASSERT_TRUE(level_files_.empty());
}
TEST(AddBoundaryInputsTest, TestEmptyCompactionFiles) {
FileMetaData* f1 =
CreateFileMetaData(1, InternalKey("100", 2, kTypeValue),
InternalKey(InternalKey("100", 1, kTypeValue)));
level_files_.push_back(f1);
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_TRUE(compaction_files_.empty());
ASSERT_EQ(1, level_files_.size());
ASSERT_EQ(f1, level_files_[0]);
}
TEST(AddBoundaryInputsTest, TestNoBoundaryFiles) {
FileMetaData* f1 =
CreateFileMetaData(1, InternalKey("100", 2, kTypeValue),
InternalKey(InternalKey("100", 1, kTypeValue)));
FileMetaData* f2 =
CreateFileMetaData(1, InternalKey("200", 2, kTypeValue),
InternalKey(InternalKey("200", 1, kTypeValue)));
FileMetaData* f3 =
CreateFileMetaData(1, InternalKey("300", 2, kTypeValue),
InternalKey(InternalKey("300", 1, kTypeValue)));
level_files_.push_back(f3);
level_files_.push_back(f2);
level_files_.push_back(f1);
compaction_files_.push_back(f2);
compaction_files_.push_back(f3);
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_EQ(2, compaction_files_.size());
}
TEST(AddBoundaryInputsTest, TestOneBoundaryFiles) {
FileMetaData* f1 =
CreateFileMetaData(1, InternalKey("100", 3, kTypeValue),
InternalKey(InternalKey("100", 2, kTypeValue)));
FileMetaData* f2 =
CreateFileMetaData(1, InternalKey("100", 1, kTypeValue),
InternalKey(InternalKey("200", 3, kTypeValue)));
FileMetaData* f3 =
CreateFileMetaData(1, InternalKey("300", 2, kTypeValue),
InternalKey(InternalKey("300", 1, kTypeValue)));
level_files_.push_back(f3);
level_files_.push_back(f2);
level_files_.push_back(f1);
compaction_files_.push_back(f1);
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_EQ(2, compaction_files_.size());
ASSERT_EQ(f1, compaction_files_[0]);
ASSERT_EQ(f2, compaction_files_[1]);
}
TEST(AddBoundaryInputsTest, TestTwoBoundaryFiles) {
FileMetaData* f1 =
CreateFileMetaData(1, InternalKey("100", 6, kTypeValue),
InternalKey(InternalKey("100", 5, kTypeValue)));
FileMetaData* f2 =
CreateFileMetaData(1, InternalKey("100", 2, kTypeValue),
InternalKey(InternalKey("300", 1, kTypeValue)));
FileMetaData* f3 =
CreateFileMetaData(1, InternalKey("100", 4, kTypeValue),
InternalKey(InternalKey("100", 3, kTypeValue)));
level_files_.push_back(f2);
level_files_.push_back(f3);
level_files_.push_back(f1);
compaction_files_.push_back(f1);
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_EQ(3, compaction_files_.size());
ASSERT_EQ(f1, compaction_files_[0]);
ASSERT_EQ(f3, compaction_files_[1]);
ASSERT_EQ(f2, compaction_files_[2]);
}
TEST(AddBoundaryInputsTest, TestDisjoinFilePointers) {
FileMetaData* f1 =
CreateFileMetaData(1, InternalKey("100", 6, kTypeValue),
InternalKey(InternalKey("100", 5, kTypeValue)));
FileMetaData* f2 =
CreateFileMetaData(1, InternalKey("100", 6, kTypeValue),
InternalKey(InternalKey("100", 5, kTypeValue)));
FileMetaData* f3 =
CreateFileMetaData(1, InternalKey("100", 2, kTypeValue),
InternalKey(InternalKey("300", 1, kTypeValue)));
FileMetaData* f4 =
CreateFileMetaData(1, InternalKey("100", 4, kTypeValue),
InternalKey(InternalKey("100", 3, kTypeValue)));
level_files_.push_back(f2);
level_files_.push_back(f3);
level_files_.push_back(f4);
compaction_files_.push_back(f1);
AddBoundaryInputs(icmp_, level_files_, &compaction_files_);
ASSERT_EQ(3, compaction_files_.size());
ASSERT_EQ(f1, compaction_files_[0]);
ASSERT_EQ(f4, compaction_files_[1]);
ASSERT_EQ(f3, compaction_files_[2]);
}
} // namespace leveldb
int main(int argc, char** argv) {
return leveldb::test::RunAllTests();
}
int main(int argc, char** argv) { return leveldb::test::RunAllTests(); }

128
issues/issue320_test.cc Normal file
View File

@@ -0,0 +1,128 @@
// Copyright (c) 2019 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include <cstdint>
#include <cstdlib>
#include <iostream>
#include <memory>
#include <string>
#include <vector>
#include "leveldb/db.h"
#include "leveldb/write_batch.h"
#include "util/testharness.h"
namespace leveldb {
namespace {
// Creates a random number in the range of [0, max).
int GenerateRandomNumber(int max) { return std::rand() % max; }
std::string CreateRandomString(int32_t index) {
static const size_t len = 1024;
char bytes[len];
size_t i = 0;
while (i < 8) {
bytes[i] = 'a' + ((index >> (4 * i)) & 0xf);
++i;
}
while (i < sizeof(bytes)) {
bytes[i] = 'a' + GenerateRandomNumber(26);
++i;
}
return std::string(bytes, sizeof(bytes));
}
} // namespace
class Issue320 {};
TEST(Issue320, Test) {
std::srand(0);
bool delete_before_put = false;
bool keep_snapshots = true;
std::vector<std::unique_ptr<std::pair<std::string, std::string>>> test_map(
10000);
std::vector<Snapshot const*> snapshots(100, nullptr);
DB* db;
Options options;
options.create_if_missing = true;
std::string dbpath = test::TmpDir() + "/leveldb_issue320_test";
ASSERT_OK(DB::Open(options, dbpath, &db));
uint32_t target_size = 10000;
uint32_t num_items = 0;
uint32_t count = 0;
std::string key;
std::string value, old_value;
WriteOptions writeOptions;
ReadOptions readOptions;
while (count < 200000) {
if ((++count % 1000) == 0) {
std::cout << "count: " << count << std::endl;
}
int index = GenerateRandomNumber(test_map.size());
WriteBatch batch;
if (test_map[index] == nullptr) {
num_items++;
test_map[index].reset(new std::pair<std::string, std::string>(
CreateRandomString(index), CreateRandomString(index)));
batch.Put(test_map[index]->first, test_map[index]->second);
} else {
ASSERT_OK(db->Get(readOptions, test_map[index]->first, &old_value));
if (old_value != test_map[index]->second) {
std::cout << "ERROR incorrect value returned by Get" << std::endl;
std::cout << " count=" << count << std::endl;
std::cout << " old value=" << old_value << std::endl;
std::cout << " test_map[index]->second=" << test_map[index]->second
<< std::endl;
std::cout << " test_map[index]->first=" << test_map[index]->first
<< std::endl;
std::cout << " index=" << index << std::endl;
ASSERT_EQ(old_value, test_map[index]->second);
}
if (num_items >= target_size && GenerateRandomNumber(100) > 30) {
batch.Delete(test_map[index]->first);
test_map[index] = nullptr;
--num_items;
} else {
test_map[index]->second = CreateRandomString(index);
if (delete_before_put) batch.Delete(test_map[index]->first);
batch.Put(test_map[index]->first, test_map[index]->second);
}
}
ASSERT_OK(db->Write(writeOptions, &batch));
if (keep_snapshots && GenerateRandomNumber(10) == 0) {
int i = GenerateRandomNumber(snapshots.size());
if (snapshots[i] != nullptr) {
db->ReleaseSnapshot(snapshots[i]);
}
snapshots[i] = db->GetSnapshot();
}
}
for (Snapshot const* snapshot : snapshots) {
if (snapshot) {
db->ReleaseSnapshot(snapshot);
}
}
delete db;
DestroyDB(dbpath, options);
}
} // namespace leveldb
int main(int argc, char** argv) { return leveldb::test::RunAllTests(); }