2016-02-09 23:12:00 +00:00
|
|
|
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
|
2017-07-15 23:03:42 +00:00
|
|
|
// 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).
|
2013-10-16 21:59:46 +00:00
|
|
|
//
|
2011-03-18 22:37:00 +00:00
|
|
|
// 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 "db/table_cache.h"
|
|
|
|
|
2015-06-23 17:25:45 +00:00
|
|
|
#include "db/dbformat.h"
|
Use only "local" range tombstones during Get (#4449)
Summary:
Previously, range tombstones were accumulated from every level, which
was necessary if a range tombstone in a higher level covered a key in a lower
level. However, RangeDelAggregator::AddTombstones's complexity is based on
the number of tombstones that are currently stored in it, which is wasteful in
the Get case, where we only need to know the highest sequence number of range
tombstones that cover the key from higher levels, and compute the highest covering
sequence number at the current level. This change introduces this optimization, and
removes the use of RangeDelAggregator from the Get path.
In the benchmark results, the following command was used to initialize the database:
```
./db_bench -db=/dev/shm/5k-rts -use_existing_db=false -benchmarks=filluniquerandom -write_buffer_size=1048576 -compression_type=lz4 -target_file_size_base=1048576 -max_bytes_for_level_base=4194304 -value_size=112 -key_size=16 -block_size=4096 -level_compaction_dynamic_level_bytes=true -num=5000000 -max_background_jobs=12 -benchmark_write_rate_limit=20971520 -range_tombstone_width=100 -writes_per_range_tombstone=100 -max_num_range_tombstones=50000 -bloom_bits=8
```
...and the following command was used to measure read throughput:
```
./db_bench -db=/dev/shm/5k-rts/ -use_existing_db=true -benchmarks=readrandom -disable_auto_compactions=true -num=5000000 -reads=100000 -threads=32
```
The filluniquerandom command was only run once, and the resulting database was used
to measure read performance before and after the PR. Both binaries were compiled with
`DEBUG_LEVEL=0`.
Readrandom results before PR:
```
readrandom : 4.544 micros/op 220090 ops/sec; 16.9 MB/s (63103 of 100000 found)
```
Readrandom results after PR:
```
readrandom : 11.147 micros/op 89707 ops/sec; 6.9 MB/s (63103 of 100000 found)
```
So it's actually slower right now, but this PR paves the way for future optimizations (see #4493).
----
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4449
Differential Revision: D10370575
Pulled By: abhimadan
fbshipit-source-id: 9a2e152be1ef36969055c0e9eb4beb0d96c11f4d
2018-10-24 19:29:29 +00:00
|
|
|
#include "db/range_tombstone_fragmenter.h"
|
2019-07-23 01:53:03 +00:00
|
|
|
#include "db/snapshot_impl.h"
|
2014-01-07 04:29:17 +00:00
|
|
|
#include "db/version_edit.h"
|
2020-06-29 21:51:57 +00:00
|
|
|
#include "file/file_util.h"
|
2019-05-30 03:44:08 +00:00
|
|
|
#include "file/filename.h"
|
2019-09-16 17:31:27 +00:00
|
|
|
#include "file/random_access_file_reader.h"
|
2017-04-06 02:02:00 +00:00
|
|
|
#include "monitoring/perf_context_imp.h"
|
2021-12-03 22:42:05 +00:00
|
|
|
#include "rocksdb/advanced_options.h"
|
2013-08-23 15:38:13 +00:00
|
|
|
#include "rocksdb/statistics.h"
|
2019-06-19 21:07:36 +00:00
|
|
|
#include "table/block_based/block_based_table_reader.h"
|
2017-04-06 02:02:00 +00:00
|
|
|
#include "table/get_context.h"
|
2015-10-12 22:06:38 +00:00
|
|
|
#include "table/internal_iterator.h"
|
In DB::NewIterator(), try to allocate the whole iterator tree in an arena
Summary:
In this patch, try to allocate the whole iterator tree starting from DBIter from an arena
1. ArenaWrappedDBIter is created when serves as the entry point of an iterator tree, with an arena in it.
2. Add an option to create iterator from arena for following iterators: DBIter, MergingIterator, MemtableIterator, all mem table's iterators, all table reader's iterators and two level iterator.
3. MergeIteratorBuilder is created to incrementally build the tree of internal iterators. It is passed to mem table list and version set and add iterators to it.
Limitations:
(1) Only DB::NewIterator() without tailing uses the arena. Other cases, including readonly DB and compactions are still from malloc
(2) Two level iterator itself is allocated in arena, but not iterators inside it.
Test Plan: make all check
Reviewers: ljin, haobo
Reviewed By: haobo
Subscribers: leveldb, dhruba, yhchiang, igor
Differential Revision: https://reviews.facebook.net/D18513
2014-06-02 23:38:00 +00:00
|
|
|
#include "table/iterator_wrapper.h"
|
Introduce a new MultiGet batching implementation (#5011)
Summary:
This PR introduces a new MultiGet() API, with the underlying implementation grouping keys based on SST file and batching lookups in a file. The reason for the new API is twofold - the definition allows callers to allocate storage for status and values on stack instead of std::vector, as well as return values as PinnableSlices in order to avoid copying, and it keeps the original MultiGet() implementation intact while we experiment with batching.
Batching is useful when there is some spatial locality to the keys being queries, as well as larger batch sizes. The main benefits are due to -
1. Fewer function calls, especially to BlockBasedTableReader::MultiGet() and FullFilterBlockReader::KeysMayMatch()
2. Bloom filter cachelines can be prefetched, hiding the cache miss latency
The next step is to optimize the binary searches in the level_storage_info, index blocks and data blocks, since we could reduce the number of key comparisons if the keys are relatively close to each other. The batching optimizations also need to be extended to other formats, such as PlainTable and filter formats. This also needs to be added to db_stress.
Benchmark results from db_bench for various batch size/locality of reference combinations are given below. Locality was simulated by offsetting the keys in a batch by a stride length. Each SST file is about 8.6MB uncompressed and key/value size is 16/100 uncompressed. To focus on the cpu benefit of batching, the runs were single threaded and bound to the same cpu to eliminate interference from other system events. The results show a 10-25% improvement in micros/op from smaller to larger batch sizes (4 - 32).
Batch Sizes
1 | 2 | 4 | 8 | 16 | 32
Random pattern (Stride length 0)
4.158 | 4.109 | 4.026 | 4.05 | 4.1 | 4.074 - Get
4.438 | 4.302 | 4.165 | 4.122 | 4.096 | 4.075 - MultiGet (no batching)
4.461 | 4.256 | 4.277 | 4.11 | 4.182 | 4.14 - MultiGet (w/ batching)
Good locality (Stride length 16)
4.048 | 3.659 | 3.248 | 2.99 | 2.84 | 2.753
4.429 | 3.728 | 3.406 | 3.053 | 2.911 | 2.781
4.452 | 3.45 | 2.833 | 2.451 | 2.233 | 2.135
Good locality (Stride length 256)
4.066 | 3.786 | 3.581 | 3.447 | 3.415 | 3.232
4.406 | 4.005 | 3.644 | 3.49 | 3.381 | 3.268
4.393 | 3.649 | 3.186 | 2.882 | 2.676 | 2.62
Medium locality (Stride length 4096)
4.012 | 3.922 | 3.768 | 3.61 | 3.582 | 3.555
4.364 | 4.057 | 3.791 | 3.65 | 3.57 | 3.465
4.479 | 3.758 | 3.316 | 3.077 | 2.959 | 2.891
dbbench command used (on a DB with 4 levels, 12 million keys)-
TEST_TMPDIR=/dev/shm numactl -C 10 ./db_bench.tmp -use_existing_db=true -benchmarks="readseq,multireadrandom" -write_buffer_size=4194304 -target_file_size_base=4194304 -max_bytes_for_level_base=16777216 -num=12000000 -reads=12000000 -duration=90 -threads=1 -compression_type=none -cache_size=4194304000 -batch_size=32 -disable_auto_compactions=true -bloom_bits=10 -cache_index_and_filter_blocks=true -pin_l0_filter_and_index_blocks_in_cache=true -multiread_batched=true -multiread_stride=4
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5011
Differential Revision: D14348703
Pulled By: anand1976
fbshipit-source-id: 774406dab3776d979c809522a67bedac6c17f84b
2019-04-11 21:24:09 +00:00
|
|
|
#include "table/multiget_context.h"
|
2015-09-11 18:36:33 +00:00
|
|
|
#include "table/table_builder.h"
|
2014-01-28 05:58:46 +00:00
|
|
|
#include "table/table_reader.h"
|
2019-05-31 00:39:43 +00:00
|
|
|
#include "test_util/sync_point.h"
|
2019-07-23 01:53:03 +00:00
|
|
|
#include "util/cast_util.h"
|
2011-03-18 22:37:00 +00:00
|
|
|
#include "util/coding.h"
|
2013-06-07 17:02:28 +00:00
|
|
|
#include "util/stop_watch.h"
|
2011-03-18 22:37:00 +00:00
|
|
|
|
2020-02-20 20:07:53 +00:00
|
|
|
namespace ROCKSDB_NAMESPACE {
|
2015-06-23 17:25:45 +00:00
|
|
|
namespace {
|
2020-03-31 23:09:11 +00:00
|
|
|
template <class T>
|
|
|
|
static void DeleteEntry(const Slice& /*key*/, void* value) {
|
|
|
|
T* typed_value = reinterpret_cast<T*>(value);
|
|
|
|
delete typed_value;
|
|
|
|
}
|
Multi file concurrency in MultiGet using coroutines and async IO (#9968)
Summary:
This PR implements a coroutine version of batched MultiGet in order to concurrently read from multiple SST files in a level using async IO, thus reducing the latency of the MultiGet. The API from the user perspective is still synchronous and single threaded, with the RocksDB part of the processing happening in the context of the caller's thread. In Version::MultiGet, the decision is made whether to call synchronous or coroutine code.
A good way to review this PR is to review the first 4 commits in order - de773b3, 70c2f70, 10b50e1, and 377a597 - before reviewing the rest.
TODO:
1. Figure out how to build it in CircleCI (requires some dependencies to be installed)
2. Do some stress testing with coroutines enabled
No regression in synchronous MultiGet between this branch and main -
```
./db_bench -use_existing_db=true --db=/data/mysql/rocksdb/prefix_scan -benchmarks="readseq,multireadrandom" -key_size=32 -value_size=512 -num=5000000 -batch_size=64 -multiread_batched=true -use_direct_reads=false -duration=60 -ops_between_duration_checks=1 -readonly=true -adaptive_readahead=true -threads=16 -cache_size=10485760000 -async_io=false -multiread_stride=40000 -statistics
```
Branch - ```multireadrandom : 4.025 micros/op 3975111 ops/sec 60.001 seconds 238509056 operations; 2062.3 MB/s (14767808 of 14767808 found)```
Main - ```multireadrandom : 3.987 micros/op 4013216 ops/sec 60.001 seconds 240795392 operations; 2082.1 MB/s (15231040 of 15231040 found)```
More benchmarks in various scenarios are given below. The measurements were taken with ```async_io=false``` (no coroutines) and ```async_io=true``` (use coroutines). For an IO bound workload (with every key requiring an IO), the coroutines version shows a clear benefit, being ~2.6X faster. For CPU bound workloads, the coroutines version has ~6-15% higher CPU utilization, depending on how many keys overlap an SST file.
1. Single thread IO bound workload on remote storage with sparse MultiGet batch keys (~1 key overlap/file) -
No coroutines - ```multireadrandom : 831.774 micros/op 1202 ops/sec 60.001 seconds 72136 operations; 0.6 MB/s (72136 of 72136 found)```
Using coroutines - ```multireadrandom : 318.742 micros/op 3137 ops/sec 60.003 seconds 188248 operations; 1.6 MB/s (188248 of 188248 found)```
2. Single thread CPU bound workload (all data cached) with ~1 key overlap/file -
No coroutines - ```multireadrandom : 4.127 micros/op 242322 ops/sec 60.000 seconds 14539384 operations; 125.7 MB/s (14539384 of 14539384 found)```
Using coroutines - ```multireadrandom : 4.741 micros/op 210935 ops/sec 60.000 seconds 12656176 operations; 109.4 MB/s (12656176 of 12656176 found)```
3. Single thread CPU bound workload with ~2 key overlap/file -
No coroutines - ```multireadrandom : 3.717 micros/op 269000 ops/sec 60.000 seconds 16140024 operations; 139.6 MB/s (16140024 of 16140024 found)```
Using coroutines - ```multireadrandom : 4.146 micros/op 241204 ops/sec 60.000 seconds 14472296 operations; 125.1 MB/s (14472296 of 14472296 found)```
4. CPU bound multi-threaded (16 threads) with ~4 key overlap/file -
No coroutines - ```multireadrandom : 4.534 micros/op 3528792 ops/sec 60.000 seconds 211728728 operations; 1830.7 MB/s (12737024 of 12737024 found) ```
Using coroutines - ```multireadrandom : 4.872 micros/op 3283812 ops/sec 60.000 seconds 197030096 operations; 1703.6 MB/s (12548032 of 12548032 found) ```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9968
Reviewed By: akankshamahajan15
Differential Revision: D36348563
Pulled By: anand1976
fbshipit-source-id: c0ce85a505fd26ebfbb09786cbd7f25202038696
2022-05-19 22:36:27 +00:00
|
|
|
} // namespace
|
|
|
|
} // namespace ROCKSDB_NAMESPACE
|
|
|
|
|
|
|
|
// Generate the regular and coroutine versions of some methods by
|
|
|
|
// including table_cache_sync_and_async.h twice
|
|
|
|
// Macros in the header will expand differently based on whether
|
|
|
|
// WITH_COROUTINES or WITHOUT_COROUTINES is defined
|
|
|
|
// clang-format off
|
|
|
|
#define WITHOUT_COROUTINES
|
|
|
|
#include "db/table_cache_sync_and_async.h"
|
|
|
|
#undef WITHOUT_COROUTINES
|
|
|
|
#define WITH_COROUTINES
|
|
|
|
#include "db/table_cache_sync_and_async.h"
|
2022-05-20 19:38:21 +00:00
|
|
|
#undef WITH_COROUTINES
|
Multi file concurrency in MultiGet using coroutines and async IO (#9968)
Summary:
This PR implements a coroutine version of batched MultiGet in order to concurrently read from multiple SST files in a level using async IO, thus reducing the latency of the MultiGet. The API from the user perspective is still synchronous and single threaded, with the RocksDB part of the processing happening in the context of the caller's thread. In Version::MultiGet, the decision is made whether to call synchronous or coroutine code.
A good way to review this PR is to review the first 4 commits in order - de773b3, 70c2f70, 10b50e1, and 377a597 - before reviewing the rest.
TODO:
1. Figure out how to build it in CircleCI (requires some dependencies to be installed)
2. Do some stress testing with coroutines enabled
No regression in synchronous MultiGet between this branch and main -
```
./db_bench -use_existing_db=true --db=/data/mysql/rocksdb/prefix_scan -benchmarks="readseq,multireadrandom" -key_size=32 -value_size=512 -num=5000000 -batch_size=64 -multiread_batched=true -use_direct_reads=false -duration=60 -ops_between_duration_checks=1 -readonly=true -adaptive_readahead=true -threads=16 -cache_size=10485760000 -async_io=false -multiread_stride=40000 -statistics
```
Branch - ```multireadrandom : 4.025 micros/op 3975111 ops/sec 60.001 seconds 238509056 operations; 2062.3 MB/s (14767808 of 14767808 found)```
Main - ```multireadrandom : 3.987 micros/op 4013216 ops/sec 60.001 seconds 240795392 operations; 2082.1 MB/s (15231040 of 15231040 found)```
More benchmarks in various scenarios are given below. The measurements were taken with ```async_io=false``` (no coroutines) and ```async_io=true``` (use coroutines). For an IO bound workload (with every key requiring an IO), the coroutines version shows a clear benefit, being ~2.6X faster. For CPU bound workloads, the coroutines version has ~6-15% higher CPU utilization, depending on how many keys overlap an SST file.
1. Single thread IO bound workload on remote storage with sparse MultiGet batch keys (~1 key overlap/file) -
No coroutines - ```multireadrandom : 831.774 micros/op 1202 ops/sec 60.001 seconds 72136 operations; 0.6 MB/s (72136 of 72136 found)```
Using coroutines - ```multireadrandom : 318.742 micros/op 3137 ops/sec 60.003 seconds 188248 operations; 1.6 MB/s (188248 of 188248 found)```
2. Single thread CPU bound workload (all data cached) with ~1 key overlap/file -
No coroutines - ```multireadrandom : 4.127 micros/op 242322 ops/sec 60.000 seconds 14539384 operations; 125.7 MB/s (14539384 of 14539384 found)```
Using coroutines - ```multireadrandom : 4.741 micros/op 210935 ops/sec 60.000 seconds 12656176 operations; 109.4 MB/s (12656176 of 12656176 found)```
3. Single thread CPU bound workload with ~2 key overlap/file -
No coroutines - ```multireadrandom : 3.717 micros/op 269000 ops/sec 60.000 seconds 16140024 operations; 139.6 MB/s (16140024 of 16140024 found)```
Using coroutines - ```multireadrandom : 4.146 micros/op 241204 ops/sec 60.000 seconds 14472296 operations; 125.1 MB/s (14472296 of 14472296 found)```
4. CPU bound multi-threaded (16 threads) with ~4 key overlap/file -
No coroutines - ```multireadrandom : 4.534 micros/op 3528792 ops/sec 60.000 seconds 211728728 operations; 1830.7 MB/s (12737024 of 12737024 found) ```
Using coroutines - ```multireadrandom : 4.872 micros/op 3283812 ops/sec 60.000 seconds 197030096 operations; 1703.6 MB/s (12548032 of 12548032 found) ```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9968
Reviewed By: akankshamahajan15
Differential Revision: D36348563
Pulled By: anand1976
fbshipit-source-id: c0ce85a505fd26ebfbb09786cbd7f25202038696
2022-05-19 22:36:27 +00:00
|
|
|
// clang-format on
|
|
|
|
|
|
|
|
namespace ROCKSDB_NAMESPACE {
|
|
|
|
|
|
|
|
namespace {
|
2020-03-31 23:09:11 +00:00
|
|
|
|
2011-03-18 22:37:00 +00:00
|
|
|
static void UnrefEntry(void* arg1, void* arg2) {
|
|
|
|
Cache* cache = reinterpret_cast<Cache*>(arg1);
|
|
|
|
Cache::Handle* h = reinterpret_cast<Cache::Handle*>(arg2);
|
|
|
|
cache->Release(h);
|
|
|
|
}
|
|
|
|
|
2014-06-13 22:54:19 +00:00
|
|
|
static Slice GetSliceForFileNumber(const uint64_t* file_number) {
|
2014-01-02 18:29:48 +00:00
|
|
|
return Slice(reinterpret_cast<const char*>(file_number),
|
|
|
|
sizeof(*file_number));
|
2013-12-27 00:25:45 +00:00
|
|
|
}
|
|
|
|
|
2015-06-23 17:25:45 +00:00
|
|
|
#ifndef ROCKSDB_LITE
|
|
|
|
|
|
|
|
void AppendVarint64(IterKey* key, uint64_t v) {
|
|
|
|
char buf[10];
|
|
|
|
auto ptr = EncodeVarint64(buf, v);
|
|
|
|
key->TrimAppend(key->Size(), buf, ptr - buf);
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif // ROCKSDB_LITE
|
|
|
|
|
|
|
|
} // namespace
|
|
|
|
|
2020-04-21 20:14:03 +00:00
|
|
|
const int kLoadConcurency = 128;
|
|
|
|
|
2021-05-05 20:59:21 +00:00
|
|
|
TableCache::TableCache(const ImmutableOptions& ioptions,
|
Fix use-after-free on implicit temporary FileOptions (#8571)
Summary:
FileOptions has an implicit conversion from EnvOptions and some
internal APIs take `const FileOptions&` and save the reference, which is
counter to Google C++ guidelines,
> Avoid defining functions that require a const reference parameter to outlive the call, because const reference parameters bind to temporaries. Instead, find a way to eliminate the lifetime requirement (for example, by copying the parameter), or pass it by const pointer and document the lifetime and non-null requirements.
This is at least a problem for repair.cc, which passes an EnvOptions to
TableCache(), which would save a reference to the temporary copy as
FileOptions. This was unfortunately only caught as a side effect of
changes in https://github.com/facebook/rocksdb/issues/8544.
This change fixes the repair.cc case and updates the involved internal
APIs that save a reference to use `const FileOptions*` instead.
Unfortunately, I don't know how to get any of our sanitizers to reliably
report bugs like this, so I can't rule out more existing in our
codebase.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8571
Test Plan:
Test that issues seen with https://github.com/facebook/rocksdb/issues/8544 are fixed (can reproduce on
AWS EC2)
Reviewed By: ajkr
Differential Revision: D29943890
Pulled By: pdillinger
fbshipit-source-id: 95f9c5251548777b4dc994c1a083dd2add5799c9
2021-07-28 04:48:22 +00:00
|
|
|
const FileOptions* file_options, Cache* const cache,
|
2020-08-27 18:20:08 +00:00
|
|
|
BlockCacheTracer* const block_cache_tracer,
|
2021-06-10 18:01:44 +00:00
|
|
|
const std::shared_ptr<IOTracer>& io_tracer,
|
|
|
|
const std::string& db_session_id)
|
2018-06-28 00:09:29 +00:00
|
|
|
: ioptions_(ioptions),
|
Fix use-after-free on implicit temporary FileOptions (#8571)
Summary:
FileOptions has an implicit conversion from EnvOptions and some
internal APIs take `const FileOptions&` and save the reference, which is
counter to Google C++ guidelines,
> Avoid defining functions that require a const reference parameter to outlive the call, because const reference parameters bind to temporaries. Instead, find a way to eliminate the lifetime requirement (for example, by copying the parameter), or pass it by const pointer and document the lifetime and non-null requirements.
This is at least a problem for repair.cc, which passes an EnvOptions to
TableCache(), which would save a reference to the temporary copy as
FileOptions. This was unfortunately only caught as a side effect of
changes in https://github.com/facebook/rocksdb/issues/8544.
This change fixes the repair.cc case and updates the involved internal
APIs that save a reference to use `const FileOptions*` instead.
Unfortunately, I don't know how to get any of our sanitizers to reliably
report bugs like this, so I can't rule out more existing in our
codebase.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8571
Test Plan:
Test that issues seen with https://github.com/facebook/rocksdb/issues/8544 are fixed (can reproduce on
AWS EC2)
Reviewed By: ajkr
Differential Revision: D29943890
Pulled By: pdillinger
fbshipit-source-id: 95f9c5251548777b4dc994c1a083dd2add5799c9
2021-07-28 04:48:22 +00:00
|
|
|
file_options_(*file_options),
|
2018-06-28 00:09:29 +00:00
|
|
|
cache_(cache),
|
2019-06-13 22:39:52 +00:00
|
|
|
immortal_tables_(false),
|
2020-04-21 20:14:03 +00:00
|
|
|
block_cache_tracer_(block_cache_tracer),
|
Integrity protection for live updates to WriteBatch (#7748)
Summary:
This PR adds the foundation classes for key-value integrity protection and the first use case: protecting live updates from the source buffers added to `WriteBatch` through the destination buffer in `MemTable`. The width of the protection info is not yet configurable -- only eight bytes per key is supported. This PR allows users to enable protection by constructing `WriteBatch` with `protection_bytes_per_key == 8`. It does not yet expose a way for users to get integrity protection via other write APIs (e.g., `Put()`, `Merge()`, `Delete()`, etc.).
The foundation classes (`ProtectionInfo.*`) embed the coverage info in their type, and provide `Protect.*()` and `Strip.*()` functions to navigate between types with different coverage. For making bytes per key configurable (for powers of two up to eight) in the future, these classes are templated on the unsigned integer type used to store the protection info. That integer contains the XOR'd result of hashes with independent seeds for all covered fields. For integer fields, the hash is computed on the raw unadjusted bytes, so the result is endian-dependent. The most significant bytes are truncated when the hash value (8 bytes) is wider than the protection integer.
When `WriteBatch` is constructed with `protection_bytes_per_key == 8`, we hold a `ProtectionInfoKVOTC` (i.e., one that covers key, value, optype aka `ValueType`, timestamp, and CF ID) for each entry added to the batch. The protection info is generated from the original buffers passed by the user, as well as the original metadata generated internally. When writing to memtable, each entry is transformed to a `ProtectionInfoKVOTS` (i.e., dropping coverage of CF ID and adding coverage of sequence number), since at that point we know the sequence number, and have already selected a memtable corresponding to a particular CF. This protection info is verified once the entry is encoded in the `MemTable` buffer.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7748
Test Plan:
- an integration test to verify a wide variety of single-byte changes to the encoded `MemTable` buffer are caught
- add to stress/crash test to verify it works in variety of configs/operations without intentional corruption
- [deferred] unit tests for `ProtectionInfo.*` classes for edge cases like KV swap, `SliceParts` and `Slice` APIs are interchangeable, etc.
Reviewed By: pdillinger
Differential Revision: D25754492
Pulled By: ajkr
fbshipit-source-id: e481bac6c03c2ab268be41359730f1ceb9964866
2021-01-29 20:17:17 +00:00
|
|
|
loader_mutex_(kLoadConcurency, kGetSliceNPHash64UnseededFnPtr),
|
2021-06-10 18:01:44 +00:00
|
|
|
io_tracer_(io_tracer),
|
|
|
|
db_session_id_(db_session_id) {
|
2015-06-23 17:25:45 +00:00
|
|
|
if (ioptions_.row_cache) {
|
|
|
|
// If the same cache is shared by multiple instances, we need to
|
|
|
|
// disambiguate its entries.
|
|
|
|
PutVarint64(&row_cache_id_, ioptions_.row_cache->NewId());
|
|
|
|
}
|
|
|
|
}
|
2011-03-18 22:37:00 +00:00
|
|
|
|
|
|
|
TableCache::~TableCache() {
|
|
|
|
}
|
|
|
|
|
2014-01-07 04:29:17 +00:00
|
|
|
TableReader* TableCache::GetTableReaderFromHandle(Cache::Handle* handle) {
|
|
|
|
return reinterpret_cast<TableReader*>(cache_->Value(handle));
|
|
|
|
}
|
|
|
|
|
|
|
|
void TableCache::ReleaseHandle(Cache::Handle* handle) {
|
|
|
|
cache_->Release(handle);
|
|
|
|
}
|
|
|
|
|
2015-08-21 05:33:25 +00:00
|
|
|
Status TableCache::GetTableReader(
|
2020-06-29 21:51:57 +00:00
|
|
|
const ReadOptions& ro, const FileOptions& file_options,
|
2015-08-21 05:33:25 +00:00
|
|
|
const InternalKeyComparator& internal_comparator, const FileDescriptor& fd,
|
2019-06-19 21:07:36 +00:00
|
|
|
bool sequential_mode, bool record_read_stats, HistogramImpl* file_read_hist,
|
|
|
|
std::unique_ptr<TableReader>* table_reader,
|
2022-01-21 19:36:36 +00:00
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor,
|
|
|
|
bool skip_filters, int level, bool prefetch_index_and_filter_in_cache,
|
2021-10-07 21:57:02 +00:00
|
|
|
size_t max_file_size_for_l0_meta_pin, Temperature file_temperature) {
|
2015-08-21 05:33:25 +00:00
|
|
|
std::string fname =
|
2018-04-06 02:49:06 +00:00
|
|
|
TableFileName(ioptions_.cf_paths, fd.GetNumber(), fd.GetPathId());
|
Introduce a new storage specific Env API (#5761)
Summary:
The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc.
This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO.
The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before.
This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection.
The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761
Differential Revision: D18868376
Pulled By: anand1976
fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
2019-12-13 22:47:08 +00:00
|
|
|
std::unique_ptr<FSRandomAccessFile> file;
|
2020-06-29 21:51:57 +00:00
|
|
|
FileOptions fopts = file_options;
|
2021-12-03 22:42:05 +00:00
|
|
|
fopts.temperature = file_temperature;
|
2021-03-15 11:32:24 +00:00
|
|
|
Status s = PrepareIOFromReadOptions(ro, ioptions_.clock, fopts.io_options);
|
2020-06-29 21:51:57 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
s = ioptions_.fs->NewRandomAccessFile(fname, fopts, &file, nullptr);
|
|
|
|
}
|
2022-03-15 16:55:49 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
RecordTick(ioptions_.stats, NO_FILE_OPENS);
|
|
|
|
} else if (s.IsPathNotFound()) {
|
2020-02-06 18:14:19 +00:00
|
|
|
fname = Rocks2LevelTableFileName(fname);
|
2021-03-15 11:32:24 +00:00
|
|
|
s = PrepareIOFromReadOptions(ro, ioptions_.clock, fopts.io_options);
|
2020-06-29 21:51:57 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
s = ioptions_.fs->NewRandomAccessFile(fname, file_options, &file,
|
|
|
|
nullptr);
|
|
|
|
}
|
2022-03-15 16:55:49 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
RecordTick(ioptions_.stats, NO_FILE_OPENS);
|
|
|
|
}
|
2020-02-06 18:14:19 +00:00
|
|
|
}
|
|
|
|
|
2015-08-21 05:33:25 +00:00
|
|
|
if (s.ok()) {
|
2015-08-26 22:25:59 +00:00
|
|
|
if (!sequential_mode && ioptions_.advise_random_on_open) {
|
Introduce a new storage specific Env API (#5761)
Summary:
The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc.
This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO.
The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before.
This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection.
The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761
Differential Revision: D18868376
Pulled By: anand1976
fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
2019-12-13 22:47:08 +00:00
|
|
|
file->Hint(FSRandomAccessFile::kRandom);
|
2015-08-21 05:33:25 +00:00
|
|
|
}
|
2021-04-26 19:43:02 +00:00
|
|
|
StopWatch sw(ioptions_.clock, ioptions_.stats, TABLE_OPEN_IO_MICROS);
|
2015-08-21 05:33:25 +00:00
|
|
|
std::unique_ptr<RandomAccessFileReader> file_reader(
|
2017-07-31 19:07:42 +00:00
|
|
|
new RandomAccessFileReader(
|
2021-03-15 11:32:24 +00:00
|
|
|
std::move(file), fname, ioptions_.clock, io_tracer_,
|
2021-04-26 19:43:02 +00:00
|
|
|
record_read_stats ? ioptions_.stats : nullptr, SST_READ_MICROS,
|
2021-10-07 21:57:02 +00:00
|
|
|
file_read_hist, ioptions_.rate_limiter.get(), ioptions_.listeners,
|
2022-02-18 21:35:36 +00:00
|
|
|
file_temperature, level == ioptions_.num_levels - 1));
|
2015-08-21 05:33:25 +00:00
|
|
|
s = ioptions_.table_factory->NewTableReader(
|
2020-06-29 21:51:57 +00:00
|
|
|
ro,
|
2021-06-10 18:01:44 +00:00
|
|
|
TableReaderOptions(
|
|
|
|
ioptions_, prefix_extractor, file_options, internal_comparator,
|
|
|
|
skip_filters, immortal_tables_, false /* force_direct_prefetch */,
|
|
|
|
level, fd.largest_seqno, block_cache_tracer_,
|
|
|
|
max_file_size_for_l0_meta_pin, db_session_id_, fd.GetNumber()),
|
2016-07-20 18:23:31 +00:00
|
|
|
std::move(file_reader), fd.GetFileSize(), table_reader,
|
|
|
|
prefetch_index_and_filter_in_cache);
|
2015-08-21 05:33:25 +00:00
|
|
|
TEST_SYNC_POINT("TableCache::GetTableReader:0");
|
|
|
|
}
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
Adding pin_l0_filter_and_index_blocks_in_cache feature and related fixes.
Summary:
When a block based table file is opened, if prefetch_index_and_filter is true, it will prefetch the index and filter blocks, putting them into the block cache.
What this feature adds: when a L0 block based table file is opened, if pin_l0_filter_and_index_blocks_in_cache is true in the options (and prefetch_index_and_filter is true), then the filter and index blocks aren't released back to the block cache at the end of BlockBasedTableReader::Open(). Instead the table reader takes ownership of them, hence pinning them, ie. the LRU cache will never push them out. Meanwhile in the table reader, further accesses will not hit the block cache, thus avoiding lock contention.
Test Plan:
'export TEST_TMPDIR=/dev/shm/ && DISABLE_JEMALLOC=1 OPT=-g make all valgrind_check -j32' is OK.
I didn't run the Java tests, I don't have Java set up on my devserver.
Reviewers: sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D56133
2016-04-01 17:42:39 +00:00
|
|
|
void TableCache::EraseHandle(const FileDescriptor& fd, Cache::Handle* handle) {
|
|
|
|
ReleaseHandle(handle);
|
|
|
|
uint64_t number = fd.GetNumber();
|
|
|
|
Slice key = GetSliceForFileNumber(&number);
|
|
|
|
cache_->Erase(key);
|
|
|
|
}
|
|
|
|
|
2021-10-07 21:57:02 +00:00
|
|
|
Status TableCache::FindTable(
|
|
|
|
const ReadOptions& ro, const FileOptions& file_options,
|
|
|
|
const InternalKeyComparator& internal_comparator, const FileDescriptor& fd,
|
2022-01-21 19:36:36 +00:00
|
|
|
Cache::Handle** handle,
|
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor,
|
2021-10-07 21:57:02 +00:00
|
|
|
const bool no_io, bool record_read_stats, HistogramImpl* file_read_hist,
|
|
|
|
bool skip_filters, int level, bool prefetch_index_and_filter_in_cache,
|
|
|
|
size_t max_file_size_for_l0_meta_pin, Temperature file_temperature) {
|
2021-03-15 11:32:24 +00:00
|
|
|
PERF_TIMER_GUARD_WITH_CLOCK(find_table_nanos, ioptions_.clock);
|
2014-07-02 16:54:20 +00:00
|
|
|
uint64_t number = fd.GetNumber();
|
|
|
|
Slice key = GetSliceForFileNumber(&number);
|
2012-04-17 15:36:46 +00:00
|
|
|
*handle = cache_->Lookup(key);
|
2015-08-21 05:33:25 +00:00
|
|
|
TEST_SYNC_POINT_CALLBACK("TableCache::FindTable:0",
|
|
|
|
const_cast<bool*>(&no_io));
|
|
|
|
|
2013-02-25 21:58:34 +00:00
|
|
|
if (*handle == nullptr) {
|
2020-09-29 23:28:42 +00:00
|
|
|
if (no_io) {
|
2013-08-25 05:48:51 +00:00
|
|
|
return Status::Incomplete("Table not found in table_cache, no_io is set");
|
2013-07-12 23:56:52 +00:00
|
|
|
}
|
2020-04-21 20:14:03 +00:00
|
|
|
MutexLock load_lock(loader_mutex_.get(key));
|
|
|
|
// We check the cache again under loading mutex
|
|
|
|
*handle = cache_->Lookup(key);
|
|
|
|
if (*handle != nullptr) {
|
2020-09-29 16:47:33 +00:00
|
|
|
return Status::OK();
|
2020-04-21 20:14:03 +00:00
|
|
|
}
|
|
|
|
|
2018-11-09 19:17:34 +00:00
|
|
|
std::unique_ptr<TableReader> table_reader;
|
2020-09-29 16:47:33 +00:00
|
|
|
Status s = GetTableReader(
|
|
|
|
ro, file_options, internal_comparator, fd, false /* sequential mode */,
|
|
|
|
record_read_stats, file_read_hist, &table_reader, prefix_extractor,
|
|
|
|
skip_filters, level, prefetch_index_and_filter_in_cache,
|
2021-10-07 21:57:02 +00:00
|
|
|
max_file_size_for_l0_meta_pin, file_temperature);
|
2011-03-18 22:37:00 +00:00
|
|
|
if (!s.ok()) {
|
2013-10-30 17:52:33 +00:00
|
|
|
assert(table_reader == nullptr);
|
2021-04-26 19:43:02 +00:00
|
|
|
RecordTick(ioptions_.stats, NO_FILE_ERRORS);
|
2011-03-18 22:37:00 +00:00
|
|
|
// We do not cache error results so that if the error is transient,
|
|
|
|
// or somebody repairs the file, we recover automatically.
|
2012-04-17 15:36:46 +00:00
|
|
|
} else {
|
2020-03-31 23:09:11 +00:00
|
|
|
s = cache_->Insert(key, table_reader.get(), 1, &DeleteEntry<TableReader>,
|
|
|
|
handle);
|
2016-03-11 01:35:19 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
// Release ownership of table reader.
|
|
|
|
table_reader.release();
|
|
|
|
}
|
2011-03-18 22:37:00 +00:00
|
|
|
}
|
2020-09-29 16:47:33 +00:00
|
|
|
return s;
|
2012-04-17 15:36:46 +00:00
|
|
|
}
|
2020-09-29 16:47:33 +00:00
|
|
|
return Status::OK();
|
2012-04-17 15:36:46 +00:00
|
|
|
}
|
2011-03-18 22:37:00 +00:00
|
|
|
|
2015-10-12 22:06:38 +00:00
|
|
|
InternalIterator* TableCache::NewIterator(
|
Introduce a new storage specific Env API (#5761)
Summary:
The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc.
This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO.
The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before.
This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection.
The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761
Differential Revision: D18868376
Pulled By: anand1976
fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
2019-12-13 22:47:08 +00:00
|
|
|
const ReadOptions& options, const FileOptions& file_options,
|
2018-07-14 00:34:54 +00:00
|
|
|
const InternalKeyComparator& icomparator, const FileMetaData& file_meta,
|
2022-01-21 19:36:36 +00:00
|
|
|
RangeDelAggregator* range_del_agg,
|
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor,
|
2018-05-21 21:33:55 +00:00
|
|
|
TableReader** table_reader_ptr, HistogramImpl* file_read_hist,
|
2019-06-20 21:28:22 +00:00
|
|
|
TableReaderCaller caller, Arena* arena, bool skip_filters, int level,
|
2020-06-09 23:49:07 +00:00
|
|
|
size_t max_file_size_for_l0_meta_pin,
|
2018-10-09 22:15:27 +00:00
|
|
|
const InternalKey* smallest_compaction_key,
|
Properly report IO errors when IndexType::kBinarySearchWithFirstKey is used (#6621)
Summary:
Context: Index type `kBinarySearchWithFirstKey` added the ability for sst file iterator to sometimes report a key from index without reading the corresponding data block. This is useful when sst blocks are cut at some meaningful boundaries (e.g. one block per key prefix), and many seeks land between blocks (e.g. for each prefix, the ranges of keys in different sst files are nearly disjoint, so a typical seek needs to read a data block from only one file even if all files have the prefix). But this added a new error condition, which rocksdb code was really not equipped to deal with: `InternalIterator::value()` may fail with an IO error or Status::Incomplete, but it's just a method returning a Slice, with no way to report error instead. Before this PR, this type of error wasn't handled at all (an empty slice was returned), and kBinarySearchWithFirstKey implementation was considered a prototype.
Now that we (LogDevice) have experimented with kBinarySearchWithFirstKey for a while and confirmed that it's really useful, this PR is adding the missing error handling.
It's a pretty inconvenient situation implementation-wise. The error needs to be reported from InternalIterator when trying to access value. But there are ~700 call sites of `InternalIterator::value()`, most of which either can't hit the error condition (because the iterator is reading from memtable or from index or something) or wouldn't benefit from the deferred loading of the value (e.g. compaction iterator that reads all values anyway). Adding error handling to all these call sites would needlessly bloat the code. So instead I made the deferred value loading optional: only the call sites that may use deferred loading have to call the new method `PrepareValue()` before calling `value()`. The feature is enabled with a new bool argument `allow_unprepared_value` to a bunch of methods that create iterators (it wouldn't make sense to put it in ReadOptions because it's completely internal to iterators, with virtually no user-visible effect). Lmk if you have better ideas.
Note that the deferred value loading only happens for *internal* iterators. The user-visible iterator (DBIter) always prepares the value before returning from Seek/Next/etc. We could go further and add an API to defer that value loading too, but that's most likely not useful for LogDevice, so it doesn't seem worth the complexity for now.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6621
Test Plan: make -j5 check . Will also deploy to some logdevice test clusters and look at stats.
Reviewed By: siying
Differential Revision: D20786930
Pulled By: al13n321
fbshipit-source-id: 6da77d918bad3780522e918f17f4d5513d3e99ee
2020-04-16 00:37:23 +00:00
|
|
|
const InternalKey* largest_compaction_key, bool allow_unprepared_value) {
|
2015-07-08 23:34:48 +00:00
|
|
|
PERF_TIMER_GUARD(new_table_iterator_nanos);
|
|
|
|
|
2016-11-17 22:30:11 +00:00
|
|
|
Status s;
|
2015-08-21 05:33:25 +00:00
|
|
|
TableReader* table_reader = nullptr;
|
2014-04-17 21:07:05 +00:00
|
|
|
Cache::Handle* handle = nullptr;
|
2017-11-03 21:34:10 +00:00
|
|
|
if (table_reader_ptr != nullptr) {
|
|
|
|
*table_reader_ptr = nullptr;
|
|
|
|
}
|
2019-06-20 21:28:22 +00:00
|
|
|
bool for_compaction = caller == TableReaderCaller::kCompaction;
|
2018-07-14 00:34:54 +00:00
|
|
|
auto& fd = file_meta.fd;
|
2019-06-19 21:07:36 +00:00
|
|
|
table_reader = fd.table_reader;
|
|
|
|
if (table_reader == nullptr) {
|
2020-06-29 21:51:57 +00:00
|
|
|
s = FindTable(
|
|
|
|
options, file_options, icomparator, fd, &handle, prefix_extractor,
|
|
|
|
options.read_tier == kBlockCacheTier /* no_io */,
|
|
|
|
!for_compaction /* record_read_stats */, file_read_hist, skip_filters,
|
|
|
|
level, true /* prefetch_index_and_filter_in_cache */,
|
2021-10-07 21:57:02 +00:00
|
|
|
max_file_size_for_l0_meta_pin, file_meta.temperature);
|
2017-11-03 21:34:10 +00:00
|
|
|
if (s.ok()) {
|
2019-06-19 21:07:36 +00:00
|
|
|
table_reader = GetTableReaderFromHandle(handle);
|
2016-11-17 22:30:11 +00:00
|
|
|
}
|
2014-01-07 04:29:17 +00:00
|
|
|
}
|
2016-11-22 05:08:06 +00:00
|
|
|
InternalIterator* result = nullptr;
|
2016-11-17 22:30:11 +00:00
|
|
|
if (s.ok()) {
|
expose a hook to skip tables during iteration
Summary:
As discussed on the mailing list (["Skipping entire SSTs while iterating"](https://groups.google.com/forum/#!topic/rocksdb/ujHCJVLrHlU)), this patch adds a `table_filter` to `ReadOptions` that allows specifying a callback to be executed during iteration before each table in the database is scanned. The callback is passed the table's properties; the table is scanned iff the callback returns true.
This can be used in conjunction with a `TablePropertiesCollector` to dramatically speed up scans by skipping tables that are known to contain irrelevant data for the scan at hand.
We're using this [downstream in CockroachDB](https://github.com/cockroachdb/cockroach/blob/master/pkg/storage/engine/db.cc#L2009-L2022) already. With this feature, under ideal conditions, we can reduce the time of an incremental backup in from hours to seconds.
FYI, the first commit in this PR fixes a segfault that I unfortunately have not figured out how to reproduce outside of CockroachDB. I'm hoping you accept it on the grounds that it is not correct to return 8-byte aligned memory from a call to `malloc` on some 64-bit platforms; one correct approach is to infer the necessary alignment from `std::max_align_t`, as done here. As noted in the first commit message, the bug is tickled by having a`std::function` in `struct ReadOptions`. That is, the following patch alone is enough to cause RocksDB to segfault when run from CockroachDB on Darwin.
```diff
--- a/include/rocksdb/options.h
+++ b/include/rocksdb/options.h
@@ -1546,6 +1546,13 @@ struct ReadOptions {
// Default: false
bool ignore_range_deletions;
+ // A callback to determine whether relevant keys for this scan exist in a
+ // given table based on the table's properties. The callback is passed the
+ // properties of each table during iteration. If the callback returns false,
+ // the table will not be scanned.
+ // Default: empty (every table will be scanned)
+ std::function<bool(const TableProperties&)> table_filter;
+
ReadOptions();
ReadOptions(bool cksum, bool cache);
};
```
/cc danhhz
Closes https://github.com/facebook/rocksdb/pull/2265
Differential Revision: D5054262
Pulled By: yiwu-arbug
fbshipit-source-id: dd6b28f2bba6cb8466250d8c5c542d3c92785476
2017-10-18 05:09:01 +00:00
|
|
|
if (options.table_filter &&
|
|
|
|
!options.table_filter(*table_reader->GetTableProperties())) {
|
2018-08-09 23:49:45 +00:00
|
|
|
result = NewEmptyInternalIterator<Slice>(arena);
|
expose a hook to skip tables during iteration
Summary:
As discussed on the mailing list (["Skipping entire SSTs while iterating"](https://groups.google.com/forum/#!topic/rocksdb/ujHCJVLrHlU)), this patch adds a `table_filter` to `ReadOptions` that allows specifying a callback to be executed during iteration before each table in the database is scanned. The callback is passed the table's properties; the table is scanned iff the callback returns true.
This can be used in conjunction with a `TablePropertiesCollector` to dramatically speed up scans by skipping tables that are known to contain irrelevant data for the scan at hand.
We're using this [downstream in CockroachDB](https://github.com/cockroachdb/cockroach/blob/master/pkg/storage/engine/db.cc#L2009-L2022) already. With this feature, under ideal conditions, we can reduce the time of an incremental backup in from hours to seconds.
FYI, the first commit in this PR fixes a segfault that I unfortunately have not figured out how to reproduce outside of CockroachDB. I'm hoping you accept it on the grounds that it is not correct to return 8-byte aligned memory from a call to `malloc` on some 64-bit platforms; one correct approach is to infer the necessary alignment from `std::max_align_t`, as done here. As noted in the first commit message, the bug is tickled by having a`std::function` in `struct ReadOptions`. That is, the following patch alone is enough to cause RocksDB to segfault when run from CockroachDB on Darwin.
```diff
--- a/include/rocksdb/options.h
+++ b/include/rocksdb/options.h
@@ -1546,6 +1546,13 @@ struct ReadOptions {
// Default: false
bool ignore_range_deletions;
+ // A callback to determine whether relevant keys for this scan exist in a
+ // given table based on the table's properties. The callback is passed the
+ // properties of each table during iteration. If the callback returns false,
+ // the table will not be scanned.
+ // Default: empty (every table will be scanned)
+ std::function<bool(const TableProperties&)> table_filter;
+
ReadOptions();
ReadOptions(bool cksum, bool cache);
};
```
/cc danhhz
Closes https://github.com/facebook/rocksdb/pull/2265
Differential Revision: D5054262
Pulled By: yiwu-arbug
fbshipit-source-id: dd6b28f2bba6cb8466250d8c5c542d3c92785476
2017-10-18 05:09:01 +00:00
|
|
|
} else {
|
2022-01-21 19:36:36 +00:00
|
|
|
result = table_reader->NewIterator(
|
|
|
|
options, prefix_extractor.get(), arena, skip_filters, caller,
|
|
|
|
file_options.compaction_readahead_size, allow_unprepared_value);
|
expose a hook to skip tables during iteration
Summary:
As discussed on the mailing list (["Skipping entire SSTs while iterating"](https://groups.google.com/forum/#!topic/rocksdb/ujHCJVLrHlU)), this patch adds a `table_filter` to `ReadOptions` that allows specifying a callback to be executed during iteration before each table in the database is scanned. The callback is passed the table's properties; the table is scanned iff the callback returns true.
This can be used in conjunction with a `TablePropertiesCollector` to dramatically speed up scans by skipping tables that are known to contain irrelevant data for the scan at hand.
We're using this [downstream in CockroachDB](https://github.com/cockroachdb/cockroach/blob/master/pkg/storage/engine/db.cc#L2009-L2022) already. With this feature, under ideal conditions, we can reduce the time of an incremental backup in from hours to seconds.
FYI, the first commit in this PR fixes a segfault that I unfortunately have not figured out how to reproduce outside of CockroachDB. I'm hoping you accept it on the grounds that it is not correct to return 8-byte aligned memory from a call to `malloc` on some 64-bit platforms; one correct approach is to infer the necessary alignment from `std::max_align_t`, as done here. As noted in the first commit message, the bug is tickled by having a`std::function` in `struct ReadOptions`. That is, the following patch alone is enough to cause RocksDB to segfault when run from CockroachDB on Darwin.
```diff
--- a/include/rocksdb/options.h
+++ b/include/rocksdb/options.h
@@ -1546,6 +1546,13 @@ struct ReadOptions {
// Default: false
bool ignore_range_deletions;
+ // A callback to determine whether relevant keys for this scan exist in a
+ // given table based on the table's properties. The callback is passed the
+ // properties of each table during iteration. If the callback returns false,
+ // the table will not be scanned.
+ // Default: empty (every table will be scanned)
+ std::function<bool(const TableProperties&)> table_filter;
+
ReadOptions();
ReadOptions(bool cksum, bool cache);
};
```
/cc danhhz
Closes https://github.com/facebook/rocksdb/pull/2265
Differential Revision: D5054262
Pulled By: yiwu-arbug
fbshipit-source-id: dd6b28f2bba6cb8466250d8c5c542d3c92785476
2017-10-18 05:09:01 +00:00
|
|
|
}
|
2019-06-19 21:07:36 +00:00
|
|
|
if (handle != nullptr) {
|
2016-11-17 22:30:11 +00:00
|
|
|
result->RegisterCleanup(&UnrefEntry, cache_, handle);
|
2016-11-22 05:08:06 +00:00
|
|
|
handle = nullptr; // prevent from releasing below
|
2016-11-17 22:30:11 +00:00
|
|
|
}
|
2013-05-17 22:53:01 +00:00
|
|
|
|
2016-11-17 22:30:11 +00:00
|
|
|
if (for_compaction) {
|
|
|
|
table_reader->SetupForCompaction();
|
|
|
|
}
|
|
|
|
if (table_reader_ptr != nullptr) {
|
|
|
|
*table_reader_ptr = table_reader;
|
|
|
|
}
|
2016-11-16 01:18:32 +00:00
|
|
|
}
|
2016-11-22 05:08:06 +00:00
|
|
|
if (s.ok() && range_del_agg != nullptr && !options.ignore_range_deletions) {
|
2018-05-04 23:37:39 +00:00
|
|
|
if (range_del_agg->AddFile(fd.GetNumber())) {
|
2018-11-21 18:53:44 +00:00
|
|
|
std::unique_ptr<FragmentedRangeTombstoneIterator> range_del_iter(
|
|
|
|
static_cast<FragmentedRangeTombstoneIterator*>(
|
|
|
|
table_reader->NewRangeTombstoneIterator(options)));
|
2018-05-04 23:37:39 +00:00
|
|
|
if (range_del_iter != nullptr) {
|
|
|
|
s = range_del_iter->status();
|
|
|
|
}
|
|
|
|
if (s.ok()) {
|
2018-10-09 22:15:27 +00:00
|
|
|
const InternalKey* smallest = &file_meta.smallest;
|
|
|
|
const InternalKey* largest = &file_meta.largest;
|
|
|
|
if (smallest_compaction_key != nullptr) {
|
|
|
|
smallest = smallest_compaction_key;
|
|
|
|
}
|
|
|
|
if (largest_compaction_key != nullptr) {
|
|
|
|
largest = largest_compaction_key;
|
|
|
|
}
|
2018-11-21 18:53:44 +00:00
|
|
|
range_del_agg->AddTombstones(std::move(range_del_iter), smallest,
|
|
|
|
largest);
|
2018-05-04 23:37:39 +00:00
|
|
|
}
|
Compaction Support for Range Deletion
Summary:
This diff introduces RangeDelAggregator, which takes ownership of iterators
provided to it via AddTombstones(). The tombstones are organized in a two-level
map (snapshot stripe -> begin key -> tombstone). Tombstone creation avoids data
copy by holding Slices returned by the iterator, which remain valid thanks to pinning.
For compaction, we create a hierarchical range tombstone iterator with structure
matching the iterator over compaction input data. An aggregator based on that
iterator is used by CompactionIterator to determine which keys are covered by
range tombstones. In case of merge operand, the same aggregator is used by
MergeHelper. Upon finishing each file in the compaction, relevant range tombstones
are added to the output file's range tombstone metablock and file boundaries are
updated accordingly.
To check whether a key is covered by range tombstone, RangeDelAggregator::ShouldDelete()
considers tombstones in the key's snapshot stripe. When this function is used outside of
compaction, it also checks newer stripes, which can contain covering tombstones. Currently
the intra-stripe check involves a linear scan; however, in the future we plan to collapse ranges
within a stripe such that binary search can be used.
RangeDelAggregator::AddToBuilder() adds all range tombstones in the table's key-range
to a new table's range tombstone meta-block. Since range tombstones may fall in the gap
between files, we may need to extend some files' key-ranges. The strategy is (1) first file
extends as far left as possible and other files do not extend left, (2) all files extend right
until either the start of the next file or the end of the last range tombstone in the gap,
whichever comes first.
One other notable change is adding release/move semantics to ScopedArenaIterator
such that it can be used to transfer ownership of an arena-allocated iterator, similar to
how unique_ptr is used for malloc'd data.
Depends on D61473
Test Plan: compaction_iterator_test, mock_table, end-to-end tests in D63927
Reviewers: sdong, IslamAbdelRahman, wanning, yhchiang, lightmark
Reviewed By: lightmark
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D62205
2016-10-18 19:04:56 +00:00
|
|
|
}
|
2016-11-16 01:18:32 +00:00
|
|
|
}
|
2016-11-22 05:08:06 +00:00
|
|
|
|
|
|
|
if (handle != nullptr) {
|
|
|
|
ReleaseHandle(handle);
|
|
|
|
}
|
|
|
|
if (!s.ok()) {
|
|
|
|
assert(result == nullptr);
|
2018-08-09 23:49:45 +00:00
|
|
|
result = NewErrorInternalIterator<Slice>(s, arena);
|
Compaction Support for Range Deletion
Summary:
This diff introduces RangeDelAggregator, which takes ownership of iterators
provided to it via AddTombstones(). The tombstones are organized in a two-level
map (snapshot stripe -> begin key -> tombstone). Tombstone creation avoids data
copy by holding Slices returned by the iterator, which remain valid thanks to pinning.
For compaction, we create a hierarchical range tombstone iterator with structure
matching the iterator over compaction input data. An aggregator based on that
iterator is used by CompactionIterator to determine which keys are covered by
range tombstones. In case of merge operand, the same aggregator is used by
MergeHelper. Upon finishing each file in the compaction, relevant range tombstones
are added to the output file's range tombstone metablock and file boundaries are
updated accordingly.
To check whether a key is covered by range tombstone, RangeDelAggregator::ShouldDelete()
considers tombstones in the key's snapshot stripe. When this function is used outside of
compaction, it also checks newer stripes, which can contain covering tombstones. Currently
the intra-stripe check involves a linear scan; however, in the future we plan to collapse ranges
within a stripe such that binary search can be used.
RangeDelAggregator::AddToBuilder() adds all range tombstones in the table's key-range
to a new table's range tombstone meta-block. Since range tombstones may fall in the gap
between files, we may need to extend some files' key-ranges. The strategy is (1) first file
extends as far left as possible and other files do not extend left, (2) all files extend right
until either the start of the next file or the end of the last range tombstone in the gap,
whichever comes first.
One other notable change is adding release/move semantics to ScopedArenaIterator
such that it can be used to transfer ownership of an arena-allocated iterator, similar to
how unique_ptr is used for malloc'd data.
Depends on D61473
Test Plan: compaction_iterator_test, mock_table, end-to-end tests in D63927
Reviewers: sdong, IslamAbdelRahman, wanning, yhchiang, lightmark
Reviewed By: lightmark
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D62205
2016-10-18 19:04:56 +00:00
|
|
|
}
|
2016-11-22 05:08:06 +00:00
|
|
|
return result;
|
2011-03-18 22:37:00 +00:00
|
|
|
}
|
|
|
|
|
2019-08-15 23:59:42 +00:00
|
|
|
Status TableCache::GetRangeTombstoneIterator(
|
|
|
|
const ReadOptions& options,
|
|
|
|
const InternalKeyComparator& internal_comparator,
|
|
|
|
const FileMetaData& file_meta,
|
|
|
|
std::unique_ptr<FragmentedRangeTombstoneIterator>* out_iter) {
|
2021-07-28 04:30:54 +00:00
|
|
|
assert(out_iter);
|
2019-08-15 23:59:42 +00:00
|
|
|
const FileDescriptor& fd = file_meta.fd;
|
|
|
|
Status s;
|
|
|
|
TableReader* t = fd.table_reader;
|
|
|
|
Cache::Handle* handle = nullptr;
|
|
|
|
if (t == nullptr) {
|
2020-06-29 21:51:57 +00:00
|
|
|
s = FindTable(options, file_options_, internal_comparator, fd, &handle);
|
2019-08-15 23:59:42 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
t = GetTableReaderFromHandle(handle);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (s.ok()) {
|
2021-07-28 04:30:54 +00:00
|
|
|
// Note: NewRangeTombstoneIterator could return nullptr
|
2019-08-15 23:59:42 +00:00
|
|
|
out_iter->reset(t->NewRangeTombstoneIterator(options));
|
2021-07-28 04:30:54 +00:00
|
|
|
}
|
|
|
|
if (handle) {
|
|
|
|
if (*out_iter) {
|
|
|
|
(*out_iter)->RegisterCleanup(&UnrefEntry, cache_, handle);
|
|
|
|
} else {
|
|
|
|
ReleaseHandle(handle);
|
|
|
|
}
|
2019-08-15 23:59:42 +00:00
|
|
|
}
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
2019-08-28 23:10:38 +00:00
|
|
|
#ifndef ROCKSDB_LITE
|
2019-09-20 19:00:55 +00:00
|
|
|
void TableCache::CreateRowCacheKeyPrefix(const ReadOptions& options,
|
|
|
|
const FileDescriptor& fd,
|
|
|
|
const Slice& internal_key,
|
|
|
|
GetContext* get_context,
|
|
|
|
IterKey& row_cache_key) {
|
2019-08-28 23:10:38 +00:00
|
|
|
uint64_t fd_number = fd.GetNumber();
|
|
|
|
// We use the user key as cache key instead of the internal key,
|
|
|
|
// otherwise the whole cache would be invalidated every time the
|
|
|
|
// sequence key increases. However, to support caching snapshot
|
|
|
|
// reads, we append the sequence number (incremented by 1 to
|
|
|
|
// distinguish from 0) only in this case.
|
|
|
|
// If the snapshot is larger than the largest seqno in the file,
|
|
|
|
// all data should be exposed to the snapshot, so we treat it
|
|
|
|
// the same as there is no snapshot. The exception is that if
|
|
|
|
// a seq-checking callback is registered, some internal keys
|
|
|
|
// may still be filtered out.
|
|
|
|
uint64_t seq_no = 0;
|
|
|
|
// Maybe we can include the whole file ifsnapshot == fd.largest_seqno.
|
|
|
|
if (options.snapshot != nullptr &&
|
|
|
|
(get_context->has_callback() ||
|
2020-04-29 20:06:27 +00:00
|
|
|
static_cast_with_check<const SnapshotImpl>(options.snapshot)
|
2019-08-28 23:10:38 +00:00
|
|
|
->GetSequenceNumber() <= fd.largest_seqno)) {
|
|
|
|
// We should consider to use options.snapshot->GetSequenceNumber()
|
|
|
|
// instead of GetInternalKeySeqno(k), which will make the code
|
|
|
|
// easier to understand.
|
|
|
|
seq_no = 1 + GetInternalKeySeqno(internal_key);
|
|
|
|
}
|
|
|
|
|
|
|
|
// Compute row cache key.
|
|
|
|
row_cache_key.TrimAppend(row_cache_key.Size(), row_cache_id_.data(),
|
|
|
|
row_cache_id_.size());
|
|
|
|
AppendVarint64(&row_cache_key, fd_number);
|
|
|
|
AppendVarint64(&row_cache_key, seq_no);
|
|
|
|
}
|
|
|
|
|
2019-09-20 19:00:55 +00:00
|
|
|
bool TableCache::GetFromRowCache(const Slice& user_key, IterKey& row_cache_key,
|
|
|
|
size_t prefix_size, GetContext* get_context) {
|
2019-08-28 23:10:38 +00:00
|
|
|
bool found = false;
|
|
|
|
|
2019-09-20 19:00:55 +00:00
|
|
|
row_cache_key.TrimAppend(prefix_size, user_key.data(), user_key.size());
|
2019-08-28 23:10:38 +00:00
|
|
|
if (auto row_handle =
|
|
|
|
ioptions_.row_cache->Lookup(row_cache_key.GetUserKey())) {
|
|
|
|
// Cleanable routine to release the cache entry
|
|
|
|
Cleanable value_pinner;
|
|
|
|
auto release_cache_entry_func = [](void* cache_to_clean,
|
|
|
|
void* cache_handle) {
|
|
|
|
((Cache*)cache_to_clean)->Release((Cache::Handle*)cache_handle);
|
|
|
|
};
|
2019-09-20 19:00:55 +00:00
|
|
|
auto found_row_cache_entry =
|
|
|
|
static_cast<const std::string*>(ioptions_.row_cache->Value(row_handle));
|
2019-08-28 23:10:38 +00:00
|
|
|
// If it comes here value is located on the cache.
|
|
|
|
// found_row_cache_entry points to the value on cache,
|
|
|
|
// and value_pinner has cleanup procedure for the cached entry.
|
|
|
|
// After replayGetContextLog() returns, get_context.pinnable_slice_
|
|
|
|
// will point to cache entry buffer (or a copy based on that) and
|
|
|
|
// cleanup routine under value_pinner will be delegated to
|
|
|
|
// get_context.pinnable_slice_. Cache entry is released when
|
|
|
|
// get_context.pinnable_slice_ is reset.
|
|
|
|
value_pinner.RegisterCleanup(release_cache_entry_func,
|
|
|
|
ioptions_.row_cache.get(), row_handle);
|
|
|
|
replayGetContextLog(*found_row_cache_entry, user_key, get_context,
|
|
|
|
&value_pinner);
|
2021-04-26 19:43:02 +00:00
|
|
|
RecordTick(ioptions_.stats, ROW_CACHE_HIT);
|
2019-08-28 23:10:38 +00:00
|
|
|
found = true;
|
|
|
|
} else {
|
2021-04-26 19:43:02 +00:00
|
|
|
RecordTick(ioptions_.stats, ROW_CACHE_MISS);
|
2019-08-28 23:10:38 +00:00
|
|
|
}
|
|
|
|
return found;
|
|
|
|
}
|
|
|
|
#endif // ROCKSDB_LITE
|
|
|
|
|
2022-01-21 19:36:36 +00:00
|
|
|
Status TableCache::Get(
|
|
|
|
const ReadOptions& options,
|
|
|
|
const InternalKeyComparator& internal_comparator,
|
|
|
|
const FileMetaData& file_meta, const Slice& k, GetContext* get_context,
|
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor,
|
|
|
|
HistogramImpl* file_read_hist, bool skip_filters, int level,
|
|
|
|
size_t max_file_size_for_l0_meta_pin) {
|
2018-07-14 00:34:54 +00:00
|
|
|
auto& fd = file_meta.fd;
|
2015-06-23 17:25:45 +00:00
|
|
|
std::string* row_cache_entry = nullptr;
|
2016-11-17 22:30:11 +00:00
|
|
|
bool done = false;
|
2015-06-23 17:25:45 +00:00
|
|
|
#ifndef ROCKSDB_LITE
|
|
|
|
IterKey row_cache_key;
|
|
|
|
std::string row_cache_entry_buffer;
|
2017-07-17 21:53:15 +00:00
|
|
|
|
2016-11-22 05:08:06 +00:00
|
|
|
// Check row cache if enabled. Since row cache does not currently store
|
|
|
|
// sequence numbers, we cannot use it if we need to fetch the sequence.
|
|
|
|
if (ioptions_.row_cache && !get_context->NeedToReadSequence()) {
|
|
|
|
auto user_key = ExtractUserKey(k);
|
2019-08-28 23:10:38 +00:00
|
|
|
CreateRowCacheKeyPrefix(options, fd, k, get_context, row_cache_key);
|
|
|
|
done = GetFromRowCache(user_key, row_cache_key, row_cache_key.Size(),
|
|
|
|
get_context);
|
|
|
|
if (!done) {
|
2016-11-22 05:08:06 +00:00
|
|
|
row_cache_entry = &row_cache_entry_buffer;
|
2015-06-23 17:25:45 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif // ROCKSDB_LITE
|
2016-11-22 05:08:06 +00:00
|
|
|
Status s;
|
|
|
|
TableReader* t = fd.table_reader;
|
|
|
|
Cache::Handle* handle = nullptr;
|
2020-09-29 16:47:33 +00:00
|
|
|
if (!done) {
|
|
|
|
assert(s.ok());
|
2016-11-22 05:08:06 +00:00
|
|
|
if (t == nullptr) {
|
2020-06-29 21:51:57 +00:00
|
|
|
s = FindTable(options, file_options_, internal_comparator, fd, &handle,
|
2020-06-09 23:49:07 +00:00
|
|
|
prefix_extractor,
|
|
|
|
options.read_tier == kBlockCacheTier /* no_io */,
|
|
|
|
true /* record_read_stats */, file_read_hist, skip_filters,
|
|
|
|
level, true /* prefetch_index_and_filter_in_cache */,
|
2021-10-07 21:57:02 +00:00
|
|
|
max_file_size_for_l0_meta_pin, file_meta.temperature);
|
2016-11-17 22:30:11 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
t = GetTableReaderFromHandle(handle);
|
|
|
|
}
|
2014-04-17 22:14:04 +00:00
|
|
|
}
|
Use only "local" range tombstones during Get (#4449)
Summary:
Previously, range tombstones were accumulated from every level, which
was necessary if a range tombstone in a higher level covered a key in a lower
level. However, RangeDelAggregator::AddTombstones's complexity is based on
the number of tombstones that are currently stored in it, which is wasteful in
the Get case, where we only need to know the highest sequence number of range
tombstones that cover the key from higher levels, and compute the highest covering
sequence number at the current level. This change introduces this optimization, and
removes the use of RangeDelAggregator from the Get path.
In the benchmark results, the following command was used to initialize the database:
```
./db_bench -db=/dev/shm/5k-rts -use_existing_db=false -benchmarks=filluniquerandom -write_buffer_size=1048576 -compression_type=lz4 -target_file_size_base=1048576 -max_bytes_for_level_base=4194304 -value_size=112 -key_size=16 -block_size=4096 -level_compaction_dynamic_level_bytes=true -num=5000000 -max_background_jobs=12 -benchmark_write_rate_limit=20971520 -range_tombstone_width=100 -writes_per_range_tombstone=100 -max_num_range_tombstones=50000 -bloom_bits=8
```
...and the following command was used to measure read throughput:
```
./db_bench -db=/dev/shm/5k-rts/ -use_existing_db=true -benchmarks=readrandom -disable_auto_compactions=true -num=5000000 -reads=100000 -threads=32
```
The filluniquerandom command was only run once, and the resulting database was used
to measure read performance before and after the PR. Both binaries were compiled with
`DEBUG_LEVEL=0`.
Readrandom results before PR:
```
readrandom : 4.544 micros/op 220090 ops/sec; 16.9 MB/s (63103 of 100000 found)
```
Readrandom results after PR:
```
readrandom : 11.147 micros/op 89707 ops/sec; 6.9 MB/s (63103 of 100000 found)
```
So it's actually slower right now, but this PR paves the way for future optimizations (see #4493).
----
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4449
Differential Revision: D10370575
Pulled By: abhimadan
fbshipit-source-id: 9a2e152be1ef36969055c0e9eb4beb0d96c11f4d
2018-10-24 19:29:29 +00:00
|
|
|
SequenceNumber* max_covering_tombstone_seq =
|
|
|
|
get_context->max_covering_tombstone_seq();
|
|
|
|
if (s.ok() && max_covering_tombstone_seq != nullptr &&
|
2016-11-29 06:45:35 +00:00
|
|
|
!options.ignore_range_deletions) {
|
2018-11-15 00:18:16 +00:00
|
|
|
std::unique_ptr<FragmentedRangeTombstoneIterator> range_del_iter(
|
2018-11-28 23:26:56 +00:00
|
|
|
t->NewRangeTombstoneIterator(options));
|
2018-11-15 00:18:16 +00:00
|
|
|
if (range_del_iter != nullptr) {
|
|
|
|
*max_covering_tombstone_seq = std::max(
|
|
|
|
*max_covering_tombstone_seq,
|
|
|
|
range_del_iter->MaxCoveringTombstoneSeqnum(ExtractUserKey(k)));
|
|
|
|
}
|
2016-11-29 06:45:35 +00:00
|
|
|
}
|
2016-11-17 22:30:11 +00:00
|
|
|
if (s.ok()) {
|
|
|
|
get_context->SetReplayLog(row_cache_entry); // nullptr if no cache.
|
2022-01-21 19:36:36 +00:00
|
|
|
s = t->Get(options, k, get_context, prefix_extractor.get(), skip_filters);
|
2016-11-17 22:30:11 +00:00
|
|
|
get_context->SetReplayLog(nullptr);
|
|
|
|
} else if (options.read_tier == kBlockCacheTier && s.IsIncomplete()) {
|
|
|
|
// Couldn't find Table in cache but treat as kFound if no_io set
|
|
|
|
get_context->MarkKeyMayExist();
|
|
|
|
s = Status::OK();
|
|
|
|
done = true;
|
2014-01-07 04:29:17 +00:00
|
|
|
}
|
2012-04-17 15:36:46 +00:00
|
|
|
}
|
2016-11-22 05:08:06 +00:00
|
|
|
|
2015-06-23 17:25:45 +00:00
|
|
|
#ifndef ROCKSDB_LITE
|
|
|
|
// Put the replay log in row cache only if something was found.
|
2016-11-17 22:30:11 +00:00
|
|
|
if (!done && s.ok() && row_cache_entry && !row_cache_entry->empty()) {
|
2022-05-09 19:27:38 +00:00
|
|
|
size_t charge = row_cache_entry->capacity() + sizeof(std::string);
|
2015-06-23 17:25:45 +00:00
|
|
|
void* row_ptr = new std::string(std::move(*row_cache_entry));
|
2020-09-29 16:47:33 +00:00
|
|
|
// If row cache is full, it's OK to continue.
|
|
|
|
ioptions_.row_cache
|
|
|
|
->Insert(row_cache_key.GetUserKey(), row_ptr, charge,
|
|
|
|
&DeleteEntry<std::string>)
|
|
|
|
.PermitUncheckedError();
|
2015-06-23 17:25:45 +00:00
|
|
|
}
|
|
|
|
#endif // ROCKSDB_LITE
|
|
|
|
|
2016-11-17 22:30:11 +00:00
|
|
|
if (handle != nullptr) {
|
|
|
|
ReleaseHandle(handle);
|
|
|
|
}
|
2012-04-17 15:36:46 +00:00
|
|
|
return s;
|
|
|
|
}
|
2014-09-04 23:18:36 +00:00
|
|
|
|
2014-02-14 00:28:21 +00:00
|
|
|
Status TableCache::GetTableProperties(
|
Introduce a new storage specific Env API (#5761)
Summary:
The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc.
This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO.
The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before.
This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection.
The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761
Differential Revision: D18868376
Pulled By: anand1976
fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
2019-12-13 22:47:08 +00:00
|
|
|
const FileOptions& file_options,
|
2014-06-13 22:54:19 +00:00
|
|
|
const InternalKeyComparator& internal_comparator, const FileDescriptor& fd,
|
2018-05-21 21:33:55 +00:00
|
|
|
std::shared_ptr<const TableProperties>* properties,
|
2022-01-21 19:36:36 +00:00
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor, bool no_io) {
|
2014-06-13 22:54:19 +00:00
|
|
|
auto table_reader = fd.table_reader;
|
2014-02-14 00:28:21 +00:00
|
|
|
// table already been pre-loaded?
|
2014-04-17 21:07:05 +00:00
|
|
|
if (table_reader) {
|
|
|
|
*properties = table_reader->GetTableProperties();
|
|
|
|
|
2020-09-29 16:47:33 +00:00
|
|
|
return Status::OK();
|
2014-02-14 00:28:21 +00:00
|
|
|
}
|
|
|
|
|
2014-04-17 21:07:05 +00:00
|
|
|
Cache::Handle* table_handle = nullptr;
|
2020-09-29 16:47:33 +00:00
|
|
|
Status s = FindTable(ReadOptions(), file_options, internal_comparator, fd,
|
|
|
|
&table_handle, prefix_extractor, no_io);
|
2014-02-14 00:28:21 +00:00
|
|
|
if (!s.ok()) {
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
assert(table_handle);
|
|
|
|
auto table = GetTableReaderFromHandle(table_handle);
|
|
|
|
*properties = table->GetTableProperties();
|
|
|
|
ReleaseHandle(table_handle);
|
|
|
|
return s;
|
|
|
|
}
|
2012-04-17 15:36:46 +00:00
|
|
|
|
2022-07-24 00:38:49 +00:00
|
|
|
Status TableCache::ApproximateKeyAnchors(
|
|
|
|
const ReadOptions& ro, const InternalKeyComparator& internal_comparator,
|
|
|
|
const FileDescriptor& fd, std::vector<TableReader::Anchor>& anchors) {
|
|
|
|
Status s;
|
|
|
|
TableReader* t = fd.table_reader;
|
|
|
|
Cache::Handle* handle = nullptr;
|
|
|
|
if (t == nullptr) {
|
|
|
|
s = FindTable(ro, file_options_, internal_comparator, fd, &handle);
|
|
|
|
if (s.ok()) {
|
|
|
|
t = GetTableReaderFromHandle(handle);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (s.ok() && t != nullptr) {
|
|
|
|
s = t->ApproximateKeyAnchors(ro, anchors);
|
|
|
|
}
|
|
|
|
if (handle != nullptr) {
|
|
|
|
ReleaseHandle(handle);
|
|
|
|
}
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
2014-08-05 18:27:34 +00:00
|
|
|
size_t TableCache::GetMemoryUsageByTableReader(
|
Introduce a new storage specific Env API (#5761)
Summary:
The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc.
This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO.
The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before.
This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection.
The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761
Differential Revision: D18868376
Pulled By: anand1976
fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
2019-12-13 22:47:08 +00:00
|
|
|
const FileOptions& file_options,
|
2018-05-21 21:33:55 +00:00
|
|
|
const InternalKeyComparator& internal_comparator, const FileDescriptor& fd,
|
2022-01-21 19:36:36 +00:00
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor) {
|
2014-08-05 18:27:34 +00:00
|
|
|
auto table_reader = fd.table_reader;
|
|
|
|
// table already been pre-loaded?
|
|
|
|
if (table_reader) {
|
|
|
|
return table_reader->ApproximateMemoryUsage();
|
|
|
|
}
|
|
|
|
|
|
|
|
Cache::Handle* table_handle = nullptr;
|
2020-10-02 00:45:52 +00:00
|
|
|
Status s = FindTable(ReadOptions(), file_options, internal_comparator, fd,
|
|
|
|
&table_handle, prefix_extractor, true);
|
2014-08-05 18:27:34 +00:00
|
|
|
if (!s.ok()) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
assert(table_handle);
|
|
|
|
auto table = GetTableReaderFromHandle(table_handle);
|
|
|
|
auto ret = table->ApproximateMemoryUsage();
|
|
|
|
ReleaseHandle(table_handle);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
Improve / clean up meta block code & integrity (#9163)
Summary:
* Checksums are now checked on meta blocks unless specifically
suppressed or not applicable (e.g. plain table). (Was other way around.)
This means a number of cases that were not checking checksums now are,
including direct read TableProperties in Version::GetTableProperties
(fixed in meta_blocks ReadTableProperties), reading any block from
PersistentCache (fixed in BlockFetcher), read TableProperties in
SstFileDumper (ldb/sst_dump/BackupEngine) before table reader open,
maybe more.
* For that to work, I moved the global_seqno+TableProperties checksum
logic to the shared table/ code, because that is used by many utilies
such as SstFileDumper.
* Also for that to work, we have to know when we're dealing with a block
that has a checksum (trailer), so added that capability to Footer based
on magic number, and from there BlockFetcher.
* Knowledge of trailer presence has also fixed a problem where other
table formats were reading blocks including bytes for a non-existant
trailer--and awkwardly kind-of not using them, e.g. no shared code
checking checksums. (BlockFetcher compression type was populated
incorrectly.) Now we only read what is needed.
* Minimized code duplication and differing/incompatible/awkward
abstractions in meta_blocks.{cc,h} (e.g. SeekTo in metaindex block
without parsing block handle)
* Moved some meta block handling code from table_properties*.*
* Moved some code specific to block-based table from shared table/ code
to BlockBasedTable class. The checksum stuff means we can't completely
separate it, but things that don't need to be in shared table/ code
should not be.
* Use unique_ptr rather than raw ptr in more places. (Note: you can
std::move from unique_ptr to shared_ptr.)
Without enhancements to GetPropertiesOfAllTablesTest (see below),
net reduction of roughly 100 lines of code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9163
Test Plan:
existing tests and
* Enhanced DBTablePropertiesTest.GetPropertiesOfAllTablesTest to verify that
checksums are now checked on direct read of table properties by TableCache
(new test would fail before this change)
* Also enhanced DBTablePropertiesTest.GetPropertiesOfAllTablesTest to test
putting table properties under old meta name
* Also generally enhanced that same test to actually test what it was
supposed to be testing already, by kicking things out of table cache when
we don't want them there.
Reviewed By: ajkr, mrambacher
Differential Revision: D32514757
Pulled By: pdillinger
fbshipit-source-id: 507964b9311d186ae8d1131182290cbd97a99fa9
2021-11-18 19:42:12 +00:00
|
|
|
bool TableCache::HasEntry(Cache* cache, uint64_t file_number) {
|
|
|
|
Cache::Handle* handle = cache->Lookup(GetSliceForFileNumber(&file_number));
|
|
|
|
if (handle) {
|
|
|
|
cache->Release(handle);
|
|
|
|
return true;
|
|
|
|
} else {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
[CF] Rethink table cache
Summary:
Adapting table cache to column families is interesting. We want table cache to be global LRU, so if some column families are use not as often as others, we want them to be evicted from cache. However, current TableCache object also constructs tables on its own. If table is not found in the cache, TableCache automatically creates new table. We want each column family to be able to specify different table factory.
To solve the problem, we still have a single LRU, but we provide the LRUCache object to TableCache on construction. We have one TableCache per column family, but the underyling cache is shared by all TableCache objects.
This allows us to have a global LRU, but still be able to support different table factories for different column families. Also, in the future it will also be able to support different directories for different column families.
Test Plan: make check
Reviewers: dhruba, haobo, kailiu, sdong
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15915
2014-02-05 17:07:55 +00:00
|
|
|
void TableCache::Evict(Cache* cache, uint64_t file_number) {
|
|
|
|
cache->Erase(GetSliceForFileNumber(&file_number));
|
2011-03-18 22:37:00 +00:00
|
|
|
}
|
|
|
|
|
2019-07-23 22:30:59 +00:00
|
|
|
uint64_t TableCache::ApproximateOffsetOf(
|
|
|
|
const Slice& key, const FileDescriptor& fd, TableReaderCaller caller,
|
|
|
|
const InternalKeyComparator& internal_comparator,
|
2022-01-21 19:36:36 +00:00
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor) {
|
2019-07-23 22:30:59 +00:00
|
|
|
uint64_t result = 0;
|
|
|
|
TableReader* table_reader = fd.table_reader;
|
|
|
|
Cache::Handle* table_handle = nullptr;
|
|
|
|
if (table_reader == nullptr) {
|
|
|
|
const bool for_compaction = (caller == TableReaderCaller::kCompaction);
|
2020-06-29 21:51:57 +00:00
|
|
|
Status s = FindTable(ReadOptions(), file_options_, internal_comparator, fd,
|
|
|
|
&table_handle, prefix_extractor, false /* no_io */,
|
2019-07-23 22:30:59 +00:00
|
|
|
!for_compaction /* record_read_stats */);
|
|
|
|
if (s.ok()) {
|
|
|
|
table_reader = GetTableReaderFromHandle(table_handle);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (table_reader != nullptr) {
|
|
|
|
result = table_reader->ApproximateOffsetOf(key, caller);
|
|
|
|
}
|
|
|
|
if (table_handle != nullptr) {
|
|
|
|
ReleaseHandle(table_handle);
|
|
|
|
}
|
|
|
|
|
|
|
|
return result;
|
|
|
|
}
|
2019-08-16 21:16:49 +00:00
|
|
|
|
|
|
|
uint64_t TableCache::ApproximateSize(
|
|
|
|
const Slice& start, const Slice& end, const FileDescriptor& fd,
|
|
|
|
TableReaderCaller caller, const InternalKeyComparator& internal_comparator,
|
2022-01-21 19:36:36 +00:00
|
|
|
const std::shared_ptr<const SliceTransform>& prefix_extractor) {
|
2019-08-16 21:16:49 +00:00
|
|
|
uint64_t result = 0;
|
|
|
|
TableReader* table_reader = fd.table_reader;
|
|
|
|
Cache::Handle* table_handle = nullptr;
|
|
|
|
if (table_reader == nullptr) {
|
|
|
|
const bool for_compaction = (caller == TableReaderCaller::kCompaction);
|
2020-06-29 21:51:57 +00:00
|
|
|
Status s = FindTable(ReadOptions(), file_options_, internal_comparator, fd,
|
|
|
|
&table_handle, prefix_extractor, false /* no_io */,
|
2019-08-16 21:16:49 +00:00
|
|
|
!for_compaction /* record_read_stats */);
|
|
|
|
if (s.ok()) {
|
|
|
|
table_reader = GetTableReaderFromHandle(table_handle);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (table_reader != nullptr) {
|
|
|
|
result = table_reader->ApproximateSize(start, end, caller);
|
|
|
|
}
|
|
|
|
if (table_handle != nullptr) {
|
|
|
|
ReleaseHandle(table_handle);
|
|
|
|
}
|
|
|
|
|
|
|
|
return result;
|
|
|
|
}
|
2020-02-20 20:07:53 +00:00
|
|
|
} // namespace ROCKSDB_NAMESPACE
|