rocksdb/util/udt_util_test.cc

508 lines
21 KiB
C++
Raw Normal View History

// Copyright (c) Meta Platforms, Inc. and affiliates.
//
// 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).
#include "util/udt_util.h"
#include <gtest/gtest.h>
#include "db/dbformat.h"
#include "test_util/testharness.h"
#include "test_util/testutil.h"
namespace ROCKSDB_NAMESPACE {
namespace {
static const std::string kTestKeyWithoutTs = "key";
static const std::string kValuePlaceHolder = "value";
Add timestamp support in dump_wal/dump/idump (#12690) Summary: As titled. For dumping wal files, since a mapping from column family id to the user comparator object is needed to print the timestamp in human readable format, option `[--db=<db_path>]` is added to `dump_wal` command to allow the user to choose to optionally open the DB as read only instance and dump the wal file with better timestamp formatting. Pull Request resolved: https://github.com/facebook/rocksdb/pull/12690 Test Plan: Manually tested dump_wal: [dump a wal file specified with --walfile] ``` >> ./ldb --walfile=$TEST_DB/000004.log dump_wal --print_value >>1,1,28,13,PUT(0) : 0x666F6F0100000000000000 : 0x7631 (Column family id: [0] contained in WAL are not opened in DB. Applied default hex formatting for user key. Specify --db=<db_path> to open DB for better user key formatting if it contains timestamp.) ``` [dump with --db specified for better timestamp formatting] ``` >> ./ldb --walfile=$TEST_DB/000004.log dump_wal --db=$TEST_DB --print_value >> 1,1,28,13,PUT(0) : 0x666F6F|timestamp:1 : 0x7631 ``` dump: [dump a file specified with --path] ``` >>./ldb --path=/tmp/rocksdbtest-501/column_family_test_75359_17910784957761284041/000004.log dump Sequence,Count,ByteSize,Physical Offset,Key(s) : value 1,1,28,13,PUT(0) : 0x666F6F0100000000000000 : 0x7631 (Column family id: [0] contained in WAL are not opened in DB. Applied default hex formatting for user key. Specify --db=<db_path> to open DB for better user key formatting if it contains timestamp.) ``` [dump db specified with --db] ``` >> ./ldb --db=/tmp/rocksdbtest-501/column_family_test_75359_17910784957761284041 dump >> foo|timestamp:1 ==> v1 Keys in range: 1 ``` idump ``` ./ldb --db=$TEST_DB idump 'foo|timestamp:1' seq:1, type:1 => v1 Internal keys in range: 1 ``` Reviewed By: ltamasi Differential Revision: D57755382 Pulled By: jowlyzhang fbshipit-source-id: a0a2ef80c92801cbf7bfccc64769c1191824362e
2024-05-24 03:26:57 +00:00
static const uint64_t kWriteUnixTime = 100;
} // namespace
class HandleTimestampSizeDifferenceTest : public testing::Test {
public:
HandleTimestampSizeDifferenceTest() = default;
// Test handler used to collect the column family id and user keys contained
// in a WriteBatch for test verification. And verifies the value part stays
// the same if it's available.
class KeyCollector : public WriteBatch::Handler {
public:
explicit KeyCollector() = default;
~KeyCollector() override = default;
Status PutCF(uint32_t cf, const Slice& key, const Slice& value) override {
if (value.compare(kValuePlaceHolder) != 0) {
return Status::InvalidArgument();
}
return AddKey(cf, key);
}
Add timestamp support in dump_wal/dump/idump (#12690) Summary: As titled. For dumping wal files, since a mapping from column family id to the user comparator object is needed to print the timestamp in human readable format, option `[--db=<db_path>]` is added to `dump_wal` command to allow the user to choose to optionally open the DB as read only instance and dump the wal file with better timestamp formatting. Pull Request resolved: https://github.com/facebook/rocksdb/pull/12690 Test Plan: Manually tested dump_wal: [dump a wal file specified with --walfile] ``` >> ./ldb --walfile=$TEST_DB/000004.log dump_wal --print_value >>1,1,28,13,PUT(0) : 0x666F6F0100000000000000 : 0x7631 (Column family id: [0] contained in WAL are not opened in DB. Applied default hex formatting for user key. Specify --db=<db_path> to open DB for better user key formatting if it contains timestamp.) ``` [dump with --db specified for better timestamp formatting] ``` >> ./ldb --walfile=$TEST_DB/000004.log dump_wal --db=$TEST_DB --print_value >> 1,1,28,13,PUT(0) : 0x666F6F|timestamp:1 : 0x7631 ``` dump: [dump a file specified with --path] ``` >>./ldb --path=/tmp/rocksdbtest-501/column_family_test_75359_17910784957761284041/000004.log dump Sequence,Count,ByteSize,Physical Offset,Key(s) : value 1,1,28,13,PUT(0) : 0x666F6F0100000000000000 : 0x7631 (Column family id: [0] contained in WAL are not opened in DB. Applied default hex formatting for user key. Specify --db=<db_path> to open DB for better user key formatting if it contains timestamp.) ``` [dump db specified with --db] ``` >> ./ldb --db=/tmp/rocksdbtest-501/column_family_test_75359_17910784957761284041 dump >> foo|timestamp:1 ==> v1 Keys in range: 1 ``` idump ``` ./ldb --db=$TEST_DB idump 'foo|timestamp:1' seq:1, type:1 => v1 Internal keys in range: 1 ``` Reviewed By: ltamasi Differential Revision: D57755382 Pulled By: jowlyzhang fbshipit-source-id: a0a2ef80c92801cbf7bfccc64769c1191824362e
2024-05-24 03:26:57 +00:00
Status TimedPutCF(uint32_t cf, const Slice& key, const Slice& value,
uint64_t write_unix_time) override {
if (value.compare(kValuePlaceHolder) != 0) {
return Status::InvalidArgument();
}
if (write_unix_time != kWriteUnixTime) {
return Status::InvalidArgument();
}
return AddKey(cf, key);
}
Status PutEntityCF(uint32_t cf, const Slice& key,
const Slice& entity) override {
Slice entity_copy = entity;
WideColumns columns;
Status s = WideColumnSerialization::Deserialize(entity_copy, columns);
if (!s.ok()) {
return s;
}
if (columns.size() != 1) {
return Status::InvalidArgument();
}
if (columns[0].value().compare(kValuePlaceHolder) != 0) {
return Status::InvalidArgument();
}
return AddKey(cf, key);
}
Status DeleteCF(uint32_t cf, const Slice& key) override {
return AddKey(cf, key);
}
Status SingleDeleteCF(uint32_t cf, const Slice& key) override {
return AddKey(cf, key);
}
Status DeleteRangeCF(uint32_t cf, const Slice& begin_key,
const Slice& end_key) override {
Status status = AddKey(cf, begin_key);
if (!status.ok()) {
return status;
}
return AddKey(cf, end_key);
}
Status MergeCF(uint32_t cf, const Slice& key, const Slice& value) override {
if (value.compare(kValuePlaceHolder) != 0) {
return Status::InvalidArgument();
}
return AddKey(cf, key);
}
Status PutBlobIndexCF(uint32_t cf, const Slice& key,
const Slice& value) override {
if (value.compare(kValuePlaceHolder) != 0) {
return Status::InvalidArgument();
}
return AddKey(cf, key);
}
Status MarkBeginPrepare(bool) override { return Status::OK(); }
Status MarkEndPrepare(const Slice&) override { return Status::OK(); }
Status MarkRollback(const Slice&) override { return Status::OK(); }
Status MarkCommit(const Slice&) override { return Status::OK(); }
Status MarkCommitWithTimestamp(const Slice&, const Slice&) override {
return Status::OK();
}
Status MarkNoop(bool) override { return Status::OK(); }
const std::vector<std::pair<uint32_t, const Slice>>& GetKeys() const {
return keys_;
}
private:
Status AddKey(uint32_t cf, const Slice& key) {
keys_.emplace_back(cf, key);
return Status::OK();
}
std::vector<std::pair<uint32_t, const Slice>> keys_;
};
void CreateKey(std::string* key_buf, size_t ts_sz) {
if (ts_sz > 0) {
AppendKeyWithMinTimestamp(key_buf, kTestKeyWithoutTs, ts_sz);
} else {
key_buf->assign(kTestKeyWithoutTs);
}
}
void CreateWriteBatch(const UnorderedMap<uint32_t, size_t>& ts_sz_for_batch,
WriteBatch* batch) {
for (const auto& [cf_id, ts_sz] : ts_sz_for_batch) {
std::string key;
CreateKey(&key, ts_sz);
ASSERT_OK(WriteBatchInternal::Put(batch, cf_id, key, kValuePlaceHolder));
ASSERT_OK(WriteBatchInternal::Delete(batch, cf_id, key));
ASSERT_OK(WriteBatchInternal::SingleDelete(batch, cf_id, key));
ASSERT_OK(WriteBatchInternal::DeleteRange(batch, cf_id, key, key));
ASSERT_OK(
WriteBatchInternal::Merge(batch, cf_id, key, kValuePlaceHolder));
ASSERT_OK(WriteBatchInternal::PutBlobIndex(batch, cf_id, key,
kValuePlaceHolder));
Add timestamp support in dump_wal/dump/idump (#12690) Summary: As titled. For dumping wal files, since a mapping from column family id to the user comparator object is needed to print the timestamp in human readable format, option `[--db=<db_path>]` is added to `dump_wal` command to allow the user to choose to optionally open the DB as read only instance and dump the wal file with better timestamp formatting. Pull Request resolved: https://github.com/facebook/rocksdb/pull/12690 Test Plan: Manually tested dump_wal: [dump a wal file specified with --walfile] ``` >> ./ldb --walfile=$TEST_DB/000004.log dump_wal --print_value >>1,1,28,13,PUT(0) : 0x666F6F0100000000000000 : 0x7631 (Column family id: [0] contained in WAL are not opened in DB. Applied default hex formatting for user key. Specify --db=<db_path> to open DB for better user key formatting if it contains timestamp.) ``` [dump with --db specified for better timestamp formatting] ``` >> ./ldb --walfile=$TEST_DB/000004.log dump_wal --db=$TEST_DB --print_value >> 1,1,28,13,PUT(0) : 0x666F6F|timestamp:1 : 0x7631 ``` dump: [dump a file specified with --path] ``` >>./ldb --path=/tmp/rocksdbtest-501/column_family_test_75359_17910784957761284041/000004.log dump Sequence,Count,ByteSize,Physical Offset,Key(s) : value 1,1,28,13,PUT(0) : 0x666F6F0100000000000000 : 0x7631 (Column family id: [0] contained in WAL are not opened in DB. Applied default hex formatting for user key. Specify --db=<db_path> to open DB for better user key formatting if it contains timestamp.) ``` [dump db specified with --db] ``` >> ./ldb --db=/tmp/rocksdbtest-501/column_family_test_75359_17910784957761284041 dump >> foo|timestamp:1 ==> v1 Keys in range: 1 ``` idump ``` ./ldb --db=$TEST_DB idump 'foo|timestamp:1' seq:1, type:1 => v1 Internal keys in range: 1 ``` Reviewed By: ltamasi Differential Revision: D57755382 Pulled By: jowlyzhang fbshipit-source-id: a0a2ef80c92801cbf7bfccc64769c1191824362e
2024-05-24 03:26:57 +00:00
ASSERT_OK(WriteBatchInternal::TimedPut(
batch, cf_id, key, kValuePlaceHolder, kWriteUnixTime));
WideColumns columns{{kDefaultWideColumnName, kValuePlaceHolder}};
ASSERT_OK(WriteBatchInternal::PutEntity(batch, cf_id, key, columns));
}
}
void CheckSequenceEqual(const WriteBatch& orig_batch,
const WriteBatch& new_batch) {
ASSERT_EQ(WriteBatchInternal::Sequence(&orig_batch),
WriteBatchInternal::Sequence(&new_batch));
}
void CheckCountEqual(const WriteBatch& orig_batch,
const WriteBatch& new_batch) {
ASSERT_EQ(WriteBatchInternal::Count(&orig_batch),
WriteBatchInternal::Count(&new_batch));
}
void VerifyKeys(
const std::vector<std::pair<uint32_t, const Slice>>& keys_with_ts,
const std::vector<std::pair<uint32_t, const Slice>>& keys_without_ts,
size_t ts_sz, std::optional<uint32_t> dropped_cf) {
ASSERT_EQ(keys_with_ts.size(), keys_without_ts.size());
const std::string kTsMin(ts_sz, static_cast<unsigned char>(0));
for (size_t i = 0; i < keys_with_ts.size(); i++) {
// TimestampRecoveryHandler ignores dropped column family and copy it over
// as is. Check the keys stay the same.
if (dropped_cf.has_value() &&
keys_with_ts[i].first == dropped_cf.value()) {
ASSERT_EQ(keys_with_ts[i].first, keys_without_ts[i].first);
ASSERT_EQ(keys_with_ts[i].second, keys_without_ts[i].second);
continue;
}
const Slice& key_with_ts = keys_with_ts[i].second;
const Slice& key_without_ts = keys_without_ts[i].second;
ASSERT_TRUE(key_with_ts.starts_with(key_without_ts));
ASSERT_EQ(key_with_ts.size() - key_without_ts.size(), ts_sz);
ASSERT_TRUE(key_with_ts.ends_with(kTsMin));
}
}
void CheckContentsWithTimestampStripping(const WriteBatch& orig_batch,
const WriteBatch& new_batch,
size_t ts_sz,
std::optional<uint32_t> dropped_cf) {
CheckSequenceEqual(orig_batch, new_batch);
CheckCountEqual(orig_batch, new_batch);
KeyCollector collector_for_orig_batch;
ASSERT_OK(orig_batch.Iterate(&collector_for_orig_batch));
KeyCollector collector_for_new_batch;
ASSERT_OK(new_batch.Iterate(&collector_for_new_batch));
VerifyKeys(collector_for_orig_batch.GetKeys(),
collector_for_new_batch.GetKeys(), ts_sz, dropped_cf);
}
void CheckContentsWithTimestampPadding(const WriteBatch& orig_batch,
const WriteBatch& new_batch,
size_t ts_sz) {
CheckSequenceEqual(orig_batch, new_batch);
CheckCountEqual(orig_batch, new_batch);
KeyCollector collector_for_orig_batch;
ASSERT_OK(orig_batch.Iterate(&collector_for_orig_batch));
KeyCollector collector_for_new_batch;
ASSERT_OK(new_batch.Iterate(&collector_for_new_batch));
VerifyKeys(collector_for_new_batch.GetKeys(),
collector_for_orig_batch.GetKeys(), ts_sz,
std::nullopt /* dropped_cf */);
}
};
TEST_F(HandleTimestampSizeDifferenceTest, AllColumnFamiliesConsistent) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{1, sizeof(uint64_t)},
{2, 0}};
UnorderedMap<uint32_t, size_t> record_ts_sz = {{1, sizeof(uint64_t)}};
WriteBatch batch;
CreateWriteBatch(running_ts_sz, &batch);
// All `check_mode` pass with OK status and `batch` not checked or updated.
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kVerifyConsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true));
std::unique_ptr<WriteBatch> new_batch(nullptr);
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true, &new_batch));
ASSERT_TRUE(new_batch.get() == nullptr);
}
TEST_F(HandleTimestampSizeDifferenceTest,
AllInconsistentColumnFamiliesDropped) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{2, 0}};
UnorderedMap<uint32_t, size_t> record_ts_sz = {{1, sizeof(uint64_t)},
{3, sizeof(char)}};
WriteBatch batch;
CreateWriteBatch(record_ts_sz, &batch);
// All `check_mode` pass with OK status and `batch` not checked or updated.
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kVerifyConsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true));
std::unique_ptr<WriteBatch> new_batch(nullptr);
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true, &new_batch));
ASSERT_TRUE(new_batch.get() == nullptr);
}
TEST_F(HandleTimestampSizeDifferenceTest, InvolvedColumnFamiliesConsistent) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{1, sizeof(uint64_t)},
{2, sizeof(char)}};
UnorderedMap<uint32_t, size_t> record_ts_sz = {{1, sizeof(uint64_t)}};
WriteBatch batch;
CreateWriteBatch(record_ts_sz, &batch);
// All `check_mode` pass with OK status and `batch` not updated.
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kVerifyConsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true));
std::unique_ptr<WriteBatch> new_batch(nullptr);
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true, &new_batch));
ASSERT_TRUE(new_batch.get() == nullptr);
}
TEST_F(HandleTimestampSizeDifferenceTest,
InconsistentColumnFamilyNeedsTimestampStripping) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{1, 0}, {2, sizeof(char)}};
UnorderedMap<uint32_t, size_t> record_ts_sz = {{1, sizeof(uint64_t)}};
WriteBatch batch;
CreateWriteBatch(record_ts_sz, &batch);
// kVerifyConsistency doesn't tolerate inconsistency for running column
// families.
ASSERT_TRUE(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kVerifyConsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true)
.IsInvalidArgument());
std::unique_ptr<WriteBatch> new_batch(nullptr);
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true, &new_batch));
ASSERT_TRUE(new_batch.get() != nullptr);
CheckContentsWithTimestampStripping(batch, *new_batch, sizeof(uint64_t),
std::nullopt /* dropped_cf */);
}
TEST_F(HandleTimestampSizeDifferenceTest,
InconsistentColumnFamilyNeedsTimestampPadding) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{1, sizeof(uint64_t)}};
// Make `record_ts_sz` not contain zero timestamp size entries to follow the
// behavior of actual WAL log timestamp size record.
UnorderedMap<uint32_t, size_t> record_ts_sz;
WriteBatch batch;
CreateWriteBatch({{1, 0}}, &batch);
// kVerifyConsistency doesn't tolerate inconsistency for running column
// families.
ASSERT_TRUE(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kVerifyConsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true)
.IsInvalidArgument());
std::unique_ptr<WriteBatch> new_batch(nullptr);
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true, &new_batch));
ASSERT_TRUE(new_batch.get() != nullptr);
CheckContentsWithTimestampPadding(batch, *new_batch, sizeof(uint64_t));
}
TEST_F(HandleTimestampSizeDifferenceTest,
InconsistencyReconcileCopyOverDroppedColumnFamily) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{1, 0}};
UnorderedMap<uint32_t, size_t> record_ts_sz = {{1, sizeof(uint64_t)},
{2, sizeof(char)}};
WriteBatch batch;
CreateWriteBatch(record_ts_sz, &batch);
std::unique_ptr<WriteBatch> new_batch(nullptr);
// kReconcileInconsistency tolerate inconsistency for dropped column family
// and all related entries copied over to the new WriteBatch.
ASSERT_OK(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true, &new_batch));
ASSERT_TRUE(new_batch.get() != nullptr);
CheckContentsWithTimestampStripping(batch, *new_batch, sizeof(uint64_t),
std::optional<uint32_t>(2));
}
TEST_F(HandleTimestampSizeDifferenceTest, UnrecoverableInconsistency) {
UnorderedMap<uint32_t, size_t> running_ts_sz = {{1, sizeof(char)}};
UnorderedMap<uint32_t, size_t> record_ts_sz = {{1, sizeof(uint64_t)}};
WriteBatch batch;
CreateWriteBatch(record_ts_sz, &batch);
ASSERT_TRUE(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kVerifyConsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true)
.IsInvalidArgument());
ASSERT_TRUE(HandleWriteBatchTimestampSizeDifference(
&batch, running_ts_sz, record_ts_sz,
TimestampSizeConsistencyMode::kReconcileInconsistency,
/* seq_per_batch */ false, /* batch_per_txn */ true)
.IsInvalidArgument());
}
Support switching on / off UDT together with in-Memtable-only feature (#11623) Summary: Add support to allow enabling / disabling user-defined timestamps feature for an existing column family in combination with the in-Memtable only feature. To do this, this PR includes: 1) Log the `persist_user_defined_timestamps` option per column family in Manifest to facilitate detecting an attempt to enable / disable UDT. This entry is enforced to be logged in the same VersionEdit as the user comparator name entry. 2) User-defined timestamps related options are validated when re-opening a column family, including user comparator name and the `persist_user_defined_timestamps` flag. These type of settings and settings change are considered valid: a) no user comparator change and no effective `persist_user_defined_timestamp` flag change. b) switch user comparator to enable UDT provided the immediately effective `persist_user_defined_timestamps` flag is false. c) switch user comparator to disable UDT provided that the before-change `persist_user_defined_timestamps` is already false. 3) when an attempt to enable UDT is detected, we mark all its existing SST files as "having no UDT" by marking its `FileMetaData.user_defined_timestamps_persisted` flag to false and handle their file boundaries `FileMetaData.smallest`, `FileMetaData.largest` by padding a min timestamp. 4) while enabling / disabling UDT feature, timestamp size inconsistency in existing WAL logs are handled to make it compatible with the running user comparator. Pull Request resolved: https://github.com/facebook/rocksdb/pull/11623 Test Plan: ``` make all check ./db_with_timestamp_basic_test --gtest-filter="*EnableDisableUDT*" ./db_wal_test --gtest_filter="*EnableDisableUDT*" ``` Reviewed By: ltamasi Differential Revision: D47636862 Pulled By: jowlyzhang fbshipit-source-id: dcd19f67292da3c3cc9584c09ad00331c9ab9322
2023-07-27 03:16:32 +00:00
TEST(ValidateUserDefinedTimestampsOptionsTest, EnableUserDefinedTimestamps) {
bool mark_sst_files = false;
const Comparator* new_comparator = test::BytewiseComparatorWithU64TsWrapper();
const Comparator* old_comparator = BytewiseComparator();
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
false /*new_persist_udt*/, true /*old_persist_udt*/, &mark_sst_files));
ASSERT_TRUE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
false /*new_persist_udt*/, false /*old_persist_udt*/, &mark_sst_files));
ASSERT_TRUE(mark_sst_files);
}
TEST(ValidateUserDefinedTimestampsOptionsTest,
EnableUserDefinedTimestampsNewPersistUDTFlagIncorrect) {
bool mark_sst_files = false;
const Comparator* new_comparator = test::BytewiseComparatorWithU64TsWrapper();
const Comparator* old_comparator = BytewiseComparator();
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
true /*new_persist_udt*/, true /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
true /*new_persist_udt*/, false /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
}
TEST(ValidateUserDefinedTimestampsOptionsTest, DisableUserDefinedTimestamps) {
bool mark_sst_files = false;
const Comparator* new_comparator = ReverseBytewiseComparator();
const Comparator* old_comparator =
test::ReverseBytewiseComparatorWithU64TsWrapper();
Support switching on / off UDT together with in-Memtable-only feature (#11623) Summary: Add support to allow enabling / disabling user-defined timestamps feature for an existing column family in combination with the in-Memtable only feature. To do this, this PR includes: 1) Log the `persist_user_defined_timestamps` option per column family in Manifest to facilitate detecting an attempt to enable / disable UDT. This entry is enforced to be logged in the same VersionEdit as the user comparator name entry. 2) User-defined timestamps related options are validated when re-opening a column family, including user comparator name and the `persist_user_defined_timestamps` flag. These type of settings and settings change are considered valid: a) no user comparator change and no effective `persist_user_defined_timestamp` flag change. b) switch user comparator to enable UDT provided the immediately effective `persist_user_defined_timestamps` flag is false. c) switch user comparator to disable UDT provided that the before-change `persist_user_defined_timestamps` is already false. 3) when an attempt to enable UDT is detected, we mark all its existing SST files as "having no UDT" by marking its `FileMetaData.user_defined_timestamps_persisted` flag to false and handle their file boundaries `FileMetaData.smallest`, `FileMetaData.largest` by padding a min timestamp. 4) while enabling / disabling UDT feature, timestamp size inconsistency in existing WAL logs are handled to make it compatible with the running user comparator. Pull Request resolved: https://github.com/facebook/rocksdb/pull/11623 Test Plan: ``` make all check ./db_with_timestamp_basic_test --gtest-filter="*EnableDisableUDT*" ./db_wal_test --gtest_filter="*EnableDisableUDT*" ``` Reviewed By: ltamasi Differential Revision: D47636862 Pulled By: jowlyzhang fbshipit-source-id: dcd19f67292da3c3cc9584c09ad00331c9ab9322
2023-07-27 03:16:32 +00:00
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
false /*new_persist_udt*/, false /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
true /*new_persist_udt*/, false /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
}
TEST(ValidateUserDefinedTimestampsOptionsTest,
DisableUserDefinedTimestampsOldPersistUDTFlagIncorrect) {
bool mark_sst_files = false;
const Comparator* new_comparator = BytewiseComparator();
const Comparator* old_comparator = test::BytewiseComparatorWithU64TsWrapper();
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
false /*new_persist_udt*/, true /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
true /*new_persist_udt*/, true /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
}
TEST(ValidateUserDefinedTimestampsOptionsTest, UserComparatorUnchanged) {
bool mark_sst_files = false;
const Comparator* ucmp_without_ts = BytewiseComparator();
const Comparator* ucmp_with_ts = test::BytewiseComparatorWithU64TsWrapper();
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
ucmp_without_ts, std::string(ucmp_without_ts->Name()),
false /*new_persist_udt*/, false /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
ucmp_without_ts, std::string(ucmp_without_ts->Name()),
true /*new_persist_udt*/, true /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
ucmp_without_ts, std::string(ucmp_without_ts->Name()),
true /*new_persist_udt*/, false /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
ucmp_without_ts, std::string(ucmp_without_ts->Name()),
false /*new_persist_udt*/, true /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
ucmp_with_ts, std::string(ucmp_with_ts->Name()), true /*new_persist_udt*/,
true /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_OK(ValidateUserDefinedTimestampsOptions(
ucmp_with_ts, std::string(ucmp_with_ts->Name()),
false /*new_persist_udt*/, false /*old_persist_udt*/, &mark_sst_files));
ASSERT_FALSE(mark_sst_files);
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
ucmp_with_ts, std::string(ucmp_with_ts->Name()),
true /*new_persist_udt*/, false /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
ucmp_with_ts, std::string(ucmp_with_ts->Name()),
false /*new_persist_udt*/, true /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
}
TEST(ValidateUserDefinedTimestampsOptionsTest, InvalidUserComparatorChange) {
bool mark_sst_files = false;
const Comparator* new_comparator = BytewiseComparator();
const Comparator* old_comparator = ReverseBytewiseComparator();
ASSERT_TRUE(ValidateUserDefinedTimestampsOptions(
new_comparator, std::string(old_comparator->Name()),
false /*new_persist_udt*/, true /*old_persist_udt*/,
&mark_sst_files)
.IsInvalidArgument());
}
TEST(GetFullHistoryTsLowFromU64CutoffTsTest, Success) {
std::string cutoff_ts;
uint64_t cutoff_ts_int = 3;
PutFixed64(&cutoff_ts, 3);
Slice cutoff_ts_slice = cutoff_ts;
std::string actual_full_history_ts_low;
GetFullHistoryTsLowFromU64CutoffTs(&cutoff_ts_slice,
&actual_full_history_ts_low);
std::string expected_ts_low;
PutFixed64(&expected_ts_low, cutoff_ts_int + 1);
ASSERT_EQ(expected_ts_low, actual_full_history_ts_low);
}
} // namespace ROCKSDB_NAMESPACE
int main(int argc, char** argv) {
ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}