rocksdb/db/blob/db_blob_index_test.cc

458 lines
16 KiB
C++
Raw Normal View History

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 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 <functional>
#include <string>
#include <utility>
#include <vector>
#include "db/arena_wrapped_db_iter.h"
#include "db/column_family.h"
#include "db/db_iter.h"
#include "db/db_test_util.h"
#include "db/dbformat.h"
#include "db/write_batch_internal.h"
#include "port/port.h"
#include "port/stack_trace.h"
#include "util/string_util.h"
#include "utilities/merge_operators.h"
namespace ROCKSDB_NAMESPACE {
// kTypeBlobIndex is a value type used by BlobDB only. The base rocksdb
// should accept the value type on write, and report not supported value
// for reads, unless caller request for it explicitly. The base rocksdb
// doesn't understand format of actual blob index (the value).
class DBBlobIndexTest : public DBTestBase {
public:
enum Tier {
kMemtable = 0,
kImmutableMemtables = 1,
kL0SstFile = 2,
kLnSstFile = 3,
};
const std::vector<Tier> kAllTiers = {Tier::kMemtable,
Tier::kImmutableMemtables,
Tier::kL0SstFile, Tier::kLnSstFile};
DBBlobIndexTest()
: DBTestBase("/db_blob_index_test", /*env_do_fsync=*/true) {}
ColumnFamilyHandle* cfh() { return dbfull()->DefaultColumnFamily(); }
ColumnFamilyData* cfd() {
return static_cast_with_check<ColumnFamilyHandleImpl>(cfh())->cfd();
}
Status PutBlobIndex(WriteBatch* batch, const Slice& key,
const Slice& blob_index) {
return WriteBatchInternal::PutBlobIndex(batch, cfd()->GetID(), key,
blob_index);
}
Status Write(WriteBatch* batch) {
return dbfull()->Write(WriteOptions(), batch);
}
std::string GetImpl(const Slice& key, bool* is_blob_index = nullptr,
const Snapshot* snapshot = nullptr) {
ReadOptions read_options;
read_options.snapshot = snapshot;
PinnableSlice value;
New API to get all merge operands for a Key (#5604) Summary: This is a new API added to db.h to allow for fetching all merge operands associated with a Key. The main motivation for this API is to support use cases where doing a full online merge is not necessary as it is performance sensitive. Example use-cases: 1. Update subset of columns and read subset of columns - Imagine a SQL Table, a row is encoded as a K/V pair (as it is done in MyRocks). If there are many columns and users only updated one of them, we can use merge operator to reduce write amplification. While users only read one or two columns in the read query, this feature can avoid a full merging of the whole row, and save some CPU. 2. Updating very few attributes in a value which is a JSON-like document - Updating one attribute can be done efficiently using merge operator, while reading back one attribute can be done more efficiently if we don't need to do a full merge. ---------------------------------------------------------------------------------------------------- API : Status GetMergeOperands( const ReadOptions& options, ColumnFamilyHandle* column_family, const Slice& key, PinnableSlice* merge_operands, GetMergeOperandsOptions* get_merge_operands_options, int* number_of_operands) Example usage : int size = 100; int number_of_operands = 0; std::vector<PinnableSlice> values(size); GetMergeOperandsOptions merge_operands_info; db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k1", values.data(), merge_operands_info, &number_of_operands); Description : Returns all the merge operands corresponding to the key. If the number of merge operands in DB is greater than merge_operands_options.expected_max_number_of_operands no merge operands are returned and status is Incomplete. Merge operands returned are in the order of insertion. merge_operands-> Points to an array of at-least merge_operands_options.expected_max_number_of_operands and the caller is responsible for allocating it. If the status returned is Incomplete then number_of_operands will contain the total number of merge operands found in DB for key. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5604 Test Plan: Added unit test and perf test in db_bench that can be run using the command: ./db_bench -benchmarks=getmergeoperands --merge_operator=sortlist Differential Revision: D16657366 Pulled By: vjnadimpalli fbshipit-source-id: 0faadd752351745224ee12d4ae9ef3cb529951bf
2019-08-06 21:22:34 +00:00
DBImpl::GetImplOptions get_impl_options;
get_impl_options.column_family = cfh();
get_impl_options.value = &value;
get_impl_options.is_blob_index = is_blob_index;
auto s = dbfull()->GetImpl(read_options, key, get_impl_options);
if (s.IsNotFound()) {
return "NOT_FOUND";
}
if (s.IsCorruption()) {
return "CORRUPTION";
}
if (s.IsNotSupported()) {
return "NOT_SUPPORTED";
}
if (!s.ok()) {
return s.ToString();
}
return value.ToString();
}
std::string GetBlobIndex(const Slice& key,
const Snapshot* snapshot = nullptr) {
bool is_blob_index = false;
std::string value = GetImpl(key, &is_blob_index, snapshot);
if (!is_blob_index) {
return "NOT_BLOB";
}
return value;
}
ArenaWrappedDBIter* GetBlobIterator() {
return dbfull()->NewIteratorImpl(
ReadOptions(), cfd(), dbfull()->GetLatestSequenceNumber(),
nullptr /*read_callback*/, true /*allow_blob*/);
}
Options GetTestOptions() {
Options options;
Fix many tests to run with MEM_ENV and ENCRYPTED_ENV; Introduce a MemoryFileSystem class (#7566) Summary: This PR does a few things: 1. The MockFileSystem class was split out from the MockEnv. This change would theoretically allow a MockFileSystem to be used by other Environments as well (if we created a means of constructing one). The MockFileSystem implements a FileSystem in its entirety and does not rely on any Wrapper implementation. 2. Make the RocksDB test suite work when MOCK_ENV=1 and ENCRYPTED_ENV=1 are set. To accomplish this, a few things were needed: - The tests that tried to use the "wrong" environment (Env::Default() instead of env_) were updated - The MockFileSystem was changed to support the features it was missing or mishandled (such as recursively deleting files in a directory or supporting renaming of a directory). 3. Updated the test framework to have a ROCKSDB_GTEST_SKIP macro. This can be used to flag tests that are skipped. Currently, this defaults to doing nothing (marks the test as SUCCESS) but will mark the tests as SKIPPED when RocksDB is upgraded to a version of gtest that supports this (gtest-1.10). I have run a full "make check" with MEM_ENV, ENCRYPTED_ENV, both, and neither under both MacOS and RedHat. A few tests were disabled/skipped for the MEM/ENCRYPTED cases. The error_handler_fs_test fails/hangs for MEM_ENV (presumably a timing problem) and I will introduce another PR/issue to track that problem. (I will also push a change to disable those tests soon). There is one more test in DBTest2 that also fails which I need to investigate or skip before this PR is merged. Theoretically, this PR should also allow the test suite to run against an Env loaded from the registry, though I do not have one to try it with currently. Finally, once this is accepted, it would be nice if there was a CircleCI job to run these tests on a checkin so this effort does not become stale. I do not know how to do that, so if someone could write that job, it would be appreciated :) Pull Request resolved: https://github.com/facebook/rocksdb/pull/7566 Reviewed By: zhichao-cao Differential Revision: D24408980 Pulled By: jay-zhuang fbshipit-source-id: 911b1554a4d0da06fd51feca0c090a4abdcb4a5f
2020-10-27 17:31:34 +00:00
options.env = CurrentOptions().env;
options.create_if_missing = true;
options.num_levels = 2;
options.disable_auto_compactions = true;
// Disable auto flushes.
options.max_write_buffer_number = 10;
options.min_write_buffer_number_to_merge = 10;
options.merge_operator = MergeOperators::CreateStringAppendOperator();
return options;
}
void MoveDataTo(Tier tier) {
switch (tier) {
case Tier::kMemtable:
break;
case Tier::kImmutableMemtables:
ASSERT_OK(dbfull()->TEST_SwitchMemtable());
break;
case Tier::kL0SstFile:
ASSERT_OK(Flush());
break;
case Tier::kLnSstFile:
ASSERT_OK(Flush());
ASSERT_OK(Put("a", "dummy"));
ASSERT_OK(Put("z", "dummy"));
ASSERT_OK(Flush());
ASSERT_OK(
dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
#ifndef ROCKSDB_LITE
ASSERT_EQ("0,1", FilesPerLevel());
#endif // !ROCKSDB_LITE
break;
}
}
};
// Should be able to write kTypeBlobIndex to memtables and SST files.
TEST_F(DBBlobIndexTest, Write) {
for (auto tier : kAllTiers) {
DestroyAndReopen(GetTestOptions());
for (int i = 1; i <= 5; i++) {
std::string index = ToString(i);
WriteBatch batch;
ASSERT_OK(PutBlobIndex(&batch, "key" + index, "blob" + index));
ASSERT_OK(Write(&batch));
}
MoveDataTo(tier);
for (int i = 1; i <= 5; i++) {
std::string index = ToString(i);
ASSERT_EQ("blob" + index, GetBlobIndex("key" + index));
}
}
}
// Note: the following test case pertains to the StackableDB-based BlobDB
// implementation. Get should be able to return blob index if is_blob_index is
// provided, otherwise it should return Status::NotSupported (when reading from
// memtable) or Status::Corruption (when reading from SST). Reading from SST
// returns Corruption because we can't differentiate between the application
// accidentally opening the base DB of a stacked BlobDB and actual corruption
// when using the integrated BlobDB.
TEST_F(DBBlobIndexTest, Get) {
for (auto tier : kAllTiers) {
DestroyAndReopen(GetTestOptions());
WriteBatch batch;
ASSERT_OK(batch.Put("key", "value"));
ASSERT_OK(PutBlobIndex(&batch, "blob_key", "blob_index"));
ASSERT_OK(Write(&batch));
MoveDataTo(tier);
// Verify normal value
bool is_blob_index = false;
PinnableSlice value;
ASSERT_EQ("value", Get("key"));
ASSERT_EQ("value", GetImpl("key"));
ASSERT_EQ("value", GetImpl("key", &is_blob_index));
ASSERT_FALSE(is_blob_index);
// Verify blob index
if (tier <= kImmutableMemtables) {
ASSERT_TRUE(Get("blob_key", &value).IsNotSupported());
ASSERT_EQ("NOT_SUPPORTED", GetImpl("blob_key"));
} else {
ASSERT_TRUE(Get("blob_key", &value).IsCorruption());
ASSERT_EQ("CORRUPTION", GetImpl("blob_key"));
}
ASSERT_EQ("blob_index", GetImpl("blob_key", &is_blob_index));
ASSERT_TRUE(is_blob_index);
}
}
// Note: the following test case pertains to the StackableDB-based BlobDB
// implementation. Get should NOT return Status::NotSupported/Status::Corruption
// if blob index is updated with a normal value. See the test case above for
// more details.
TEST_F(DBBlobIndexTest, Updated) {
for (auto tier : kAllTiers) {
DestroyAndReopen(GetTestOptions());
WriteBatch batch;
for (int i = 0; i < 10; i++) {
ASSERT_OK(PutBlobIndex(&batch, "key" + ToString(i), "blob_index"));
}
ASSERT_OK(Write(&batch));
// Avoid blob values from being purged.
const Snapshot* snapshot = dbfull()->GetSnapshot();
ASSERT_OK(Put("key1", "new_value"));
ASSERT_OK(Merge("key2", "a"));
ASSERT_OK(Merge("key2", "b"));
ASSERT_OK(Merge("key2", "c"));
ASSERT_OK(Delete("key3"));
ASSERT_OK(SingleDelete("key4"));
ASSERT_OK(Delete("key5"));
ASSERT_OK(Merge("key5", "a"));
ASSERT_OK(Merge("key5", "b"));
ASSERT_OK(Merge("key5", "c"));
ASSERT_OK(dbfull()->DeleteRange(WriteOptions(), cfh(), "key6", "key9"));
MoveDataTo(tier);
for (int i = 0; i < 10; i++) {
ASSERT_EQ("blob_index", GetBlobIndex("key" + ToString(i), snapshot));
}
ASSERT_EQ("new_value", Get("key1"));
if (tier <= kImmutableMemtables) {
ASSERT_EQ("NOT_SUPPORTED", GetImpl("key2"));
} else {
ASSERT_EQ("CORRUPTION", GetImpl("key2"));
}
ASSERT_EQ("NOT_FOUND", Get("key3"));
ASSERT_EQ("NOT_FOUND", Get("key4"));
ASSERT_EQ("a,b,c", GetImpl("key5"));
for (int i = 6; i < 9; i++) {
ASSERT_EQ("NOT_FOUND", Get("key" + ToString(i)));
}
ASSERT_EQ("blob_index", GetBlobIndex("key9"));
dbfull()->ReleaseSnapshot(snapshot);
}
}
// Iterator should get blob value if allow_blob flag is set,
// otherwise return Status::NotSupported status.
TEST_F(DBBlobIndexTest, Iterate) {
const std::vector<std::vector<ValueType>> data = {
/*00*/ {kTypeValue},
/*01*/ {kTypeBlobIndex},
/*02*/ {kTypeValue},
/*03*/ {kTypeBlobIndex, kTypeValue},
/*04*/ {kTypeValue},
/*05*/ {kTypeValue, kTypeBlobIndex},
/*06*/ {kTypeValue},
/*07*/ {kTypeDeletion, kTypeBlobIndex},
/*08*/ {kTypeValue},
/*09*/ {kTypeSingleDeletion, kTypeBlobIndex},
/*10*/ {kTypeValue},
/*11*/ {kTypeMerge, kTypeMerge, kTypeMerge, kTypeBlobIndex},
/*12*/ {kTypeValue},
/*13*/
{kTypeMerge, kTypeMerge, kTypeMerge, kTypeDeletion, kTypeBlobIndex},
/*14*/ {kTypeValue},
/*15*/ {kTypeBlobIndex},
/*16*/ {kTypeValue},
};
auto get_key = [](int index) {
char buf[20];
snprintf(buf, sizeof(buf), "%02d", index);
return "key" + std::string(buf);
};
auto get_value = [&](int index, int version) {
return get_key(index) + "_value" + ToString(version);
};
auto check_iterator = [&](Iterator* iterator, Status::Code expected_status,
const Slice& expected_value) {
ASSERT_EQ(expected_status, iterator->status().code());
if (expected_status == Status::kOk) {
ASSERT_TRUE(iterator->Valid());
ASSERT_EQ(expected_value, iterator->value());
} else {
ASSERT_FALSE(iterator->Valid());
}
};
auto create_normal_iterator = [&]() -> Iterator* {
return dbfull()->NewIterator(ReadOptions());
};
auto create_blob_iterator = [&]() -> Iterator* { return GetBlobIterator(); };
auto check_is_blob = [&](bool is_blob) {
return [is_blob](Iterator* iterator) {
ASSERT_EQ(is_blob,
reinterpret_cast<ArenaWrappedDBIter*>(iterator)->IsBlob());
};
};
auto verify = [&](int index, Status::Code expected_status,
const Slice& forward_value, const Slice& backward_value,
std::function<Iterator*()> create_iterator,
std::function<void(Iterator*)> extra_check = nullptr) {
// Seek
auto* iterator = create_iterator();
ASSERT_OK(iterator->Refresh());
iterator->Seek(get_key(index));
check_iterator(iterator, expected_status, forward_value);
if (extra_check) {
extra_check(iterator);
}
delete iterator;
// Next
iterator = create_iterator();
ASSERT_OK(iterator->Refresh());
iterator->Seek(get_key(index - 1));
ASSERT_TRUE(iterator->Valid());
iterator->Next();
check_iterator(iterator, expected_status, forward_value);
if (extra_check) {
extra_check(iterator);
}
delete iterator;
// SeekForPrev
iterator = create_iterator();
ASSERT_OK(iterator->Refresh());
iterator->SeekForPrev(get_key(index));
check_iterator(iterator, expected_status, backward_value);
if (extra_check) {
extra_check(iterator);
}
delete iterator;
// Prev
iterator = create_iterator();
iterator->Seek(get_key(index + 1));
ASSERT_TRUE(iterator->Valid());
iterator->Prev();
check_iterator(iterator, expected_status, backward_value);
if (extra_check) {
extra_check(iterator);
}
delete iterator;
};
for (auto tier : {Tier::kMemtable} /*kAllTiers*/) {
// Avoid values from being purged.
std::vector<const Snapshot*> snapshots;
DestroyAndReopen(GetTestOptions());
// fill data
for (int i = 0; i < static_cast<int>(data.size()); i++) {
for (int j = static_cast<int>(data[i].size()) - 1; j >= 0; j--) {
std::string key = get_key(i);
std::string value = get_value(i, j);
WriteBatch batch;
switch (data[i][j]) {
case kTypeValue:
ASSERT_OK(Put(key, value));
break;
case kTypeDeletion:
ASSERT_OK(Delete(key));
break;
case kTypeSingleDeletion:
ASSERT_OK(SingleDelete(key));
break;
case kTypeMerge:
ASSERT_OK(Merge(key, value));
break;
case kTypeBlobIndex:
ASSERT_OK(PutBlobIndex(&batch, key, value));
ASSERT_OK(Write(&batch));
break;
default:
assert(false);
};
}
snapshots.push_back(dbfull()->GetSnapshot());
}
ASSERT_OK(
dbfull()->DeleteRange(WriteOptions(), cfh(), get_key(15), get_key(16)));
snapshots.push_back(dbfull()->GetSnapshot());
MoveDataTo(tier);
// Normal iterator
verify(1, Status::kNotSupported, "", "", create_normal_iterator);
verify(3, Status::kNotSupported, "", "", create_normal_iterator);
verify(5, Status::kOk, get_value(5, 0), get_value(5, 0),
create_normal_iterator);
verify(7, Status::kOk, get_value(8, 0), get_value(6, 0),
create_normal_iterator);
verify(9, Status::kOk, get_value(10, 0), get_value(8, 0),
create_normal_iterator);
verify(11, Status::kNotSupported, "", "", create_normal_iterator);
verify(13, Status::kOk,
get_value(13, 2) + "," + get_value(13, 1) + "," + get_value(13, 0),
get_value(13, 2) + "," + get_value(13, 1) + "," + get_value(13, 0),
create_normal_iterator);
verify(15, Status::kOk, get_value(16, 0), get_value(14, 0),
create_normal_iterator);
// Iterator with blob support
verify(1, Status::kOk, get_value(1, 0), get_value(1, 0),
create_blob_iterator, check_is_blob(true));
verify(3, Status::kOk, get_value(3, 0), get_value(3, 0),
create_blob_iterator, check_is_blob(true));
verify(5, Status::kOk, get_value(5, 0), get_value(5, 0),
create_blob_iterator, check_is_blob(false));
verify(7, Status::kOk, get_value(8, 0), get_value(6, 0),
create_blob_iterator, check_is_blob(false));
verify(9, Status::kOk, get_value(10, 0), get_value(8, 0),
create_blob_iterator, check_is_blob(false));
verify(11, Status::kNotSupported, "", "", create_blob_iterator);
verify(13, Status::kOk,
get_value(13, 2) + "," + get_value(13, 1) + "," + get_value(13, 0),
get_value(13, 2) + "," + get_value(13, 1) + "," + get_value(13, 0),
create_blob_iterator, check_is_blob(false));
verify(15, Status::kOk, get_value(16, 0), get_value(14, 0),
create_blob_iterator, check_is_blob(false));
#ifndef ROCKSDB_LITE
// Iterator with blob support and using seek.
ASSERT_OK(dbfull()->SetOptions(
cfh(), {{"max_sequential_skip_in_iterations", "0"}}));
verify(1, Status::kOk, get_value(1, 0), get_value(1, 0),
create_blob_iterator, check_is_blob(true));
verify(3, Status::kOk, get_value(3, 0), get_value(3, 0),
create_blob_iterator, check_is_blob(true));
verify(5, Status::kOk, get_value(5, 0), get_value(5, 0),
create_blob_iterator, check_is_blob(false));
verify(7, Status::kOk, get_value(8, 0), get_value(6, 0),
create_blob_iterator, check_is_blob(false));
verify(9, Status::kOk, get_value(10, 0), get_value(8, 0),
create_blob_iterator, check_is_blob(false));
verify(11, Status::kNotSupported, "", "", create_blob_iterator);
verify(13, Status::kOk,
get_value(13, 2) + "," + get_value(13, 1) + "," + get_value(13, 0),
get_value(13, 2) + "," + get_value(13, 1) + "," + get_value(13, 0),
create_blob_iterator, check_is_blob(false));
verify(15, Status::kOk, get_value(16, 0), get_value(14, 0),
create_blob_iterator, check_is_blob(false));
#endif // !ROCKSDB_LITE
for (auto* snapshot : snapshots) {
dbfull()->ReleaseSnapshot(snapshot);
}
}
}
} // namespace ROCKSDB_NAMESPACE
int main(int argc, char** argv) {
ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}