rocksdb/db/table_cache.h
Peter Dillinger 9f7801c5f1 Major Cache refactoring, CPU efficiency improvement (#10975)
Summary:
This is several refactorings bundled into one to avoid having to incrementally re-modify uses of Cache several times. Overall, there are breaking changes to Cache class, and it becomes more of low-level interface for implementing caches, especially block cache. New internal APIs make using Cache cleaner than before, and more insulated from block cache evolution. Hopefully, this is the last really big block cache refactoring, because of rather effectively decoupling the implementations from the uses. This change also removes the EXPERIMENTAL designation on the SecondaryCache support in Cache. It seems reasonably mature at this point but still subject to change/evolution (as I warn in the API docs for Cache).

The high-level motivation for this refactoring is to minimize code duplication / compounding complexity in adding SecondaryCache support to HyperClockCache (in a later PR). Other benefits listed below.

* static_cast lines of code +29 -35 (net removed 6)
* reinterpret_cast lines of code +6 -32 (net removed 26)

## cache.h and secondary_cache.h
* Always use CacheItemHelper with entries instead of just a Deleter. There are several motivations / justifications:
  * Simpler for implementations to deal with just one Insert and one Lookup.
  * Simpler and more efficient implementation because we don't have to track which entries are using helpers and which are using deleters
  * Gets rid of hack to classify cache entries by their deleter. Instead, the CacheItemHelper includes a CacheEntryRole. This simplifies a lot of code (cache_entry_roles.h almost eliminated). Fixes https://github.com/facebook/rocksdb/issues/9428.
  * Makes it trivial to adjust SecondaryCache behavior based on kind of block (e.g. don't re-compress filter blocks).
  * It is arguably less convenient for many direct users of Cache, but direct users of Cache are now rare with introduction of typed_cache.h (below).
  * I considered and rejected an alternative approach in which we reduce customizability by assuming each secondary cache compatible value starts with a Slice referencing the uncompressed block contents (already true or mostly true), but we apparently intend to stack secondary caches. Saving an entry from a compressed secondary to a lower tier requires custom handling offered by SaveToCallback, etc.
* Make CreateCallback part of the helper and introduce CreateContext to work with it (alternative to https://github.com/facebook/rocksdb/issues/10562). This cleans up the interface while still allowing context to be provided for loading/parsing values into primary cache. This model works for async lookup in BlockBasedTable reader (reader owns a CreateContext) under the assumption that it always waits on secondary cache operations to finish. (Otherwise, the CreateContext could be destroyed while async operation depending on it continues.) This likely contributes most to the observed performance improvement because it saves an std::function backed by a heap allocation.
* Use char* for serialized data, e.g. in SaveToCallback, where void* was confusingly used. (We use `char*` for serialized byte data all over RocksDB, with many advantages over `void*`. `memcpy` etc. are legacy APIs that should not be mimicked.)
* Add a type alias Cache::ObjectPtr = void*, so that we can better indicate the intent of the void* when it is to be the object associated with a Cache entry. Related: started (but did not complete) a refactoring to move away from "value" of a cache entry toward "object" or "obj". (It is confusing to call Cache a key-value store (like DB) when it is really storing arbitrary in-memory objects, not byte strings.)
* Remove unnecessary key param from DeleterFn. This is good for efficiency in HyperClockCache, which does not directly store the cache key in memory. (Alternative to https://github.com/facebook/rocksdb/issues/10774)
* Add allocator to Cache DeleterFn. This is a kind of future-proofing change in case we get more serious about using the Cache allocator for memory tracked by the Cache. Right now, only the uncompressed block contents are allocated using the allocator, and a pointer to that allocator is saved as part of the cached object so that the deleter can use it. (See CacheAllocationPtr.) If in the future we are able to "flatten out" our Cache objects some more, it would be good not to have to track the allocator as part of each object.
* Removes legacy `ApplyToAllCacheEntries` and changes `ApplyToAllEntries` signature for Deleter->CacheItemHelper change.

## typed_cache.h
Adds various "typed" interfaces to the Cache as internal APIs, so that most uses of Cache can use simple type safe code without casting and without explicit deleters, etc. Almost all of the non-test, non-glue code uses of Cache have been migrated. (Follow-up work: CompressedSecondaryCache deserves deeper attention to migrate.) This change expands RocksDB's internal usage of metaprogramming and SFINAE (https://en.cppreference.com/w/cpp/language/sfinae).

The existing usages of Cache are divided up at a high level into these new interfaces. See updated existing uses of Cache for examples of how these are used.
* PlaceholderCacheInterface - Used for making cache reservations, with entries that have a charge but no value.
* BasicTypedCacheInterface<TValue> - Used for primary cache storage of objects of type TValue, which can be cleaned up with std::default_delete<TValue>. The role is provided by TValue::kCacheEntryRole or given in an optional template parameter.
* FullTypedCacheInterface<TValue, TCreateContext> - Used for secondary cache compatible storage of objects of type TValue. In addition to BasicTypedCacheInterface constraints, we require TValue::ContentSlice() to return persistable data. This simplifies usage for the normal case of simple secondary cache compatibility (can give you a Slice to the data already in memory). In addition to TCreateContext performing the role of Cache::CreateContext, it is also expected to provide a factory function for creating TValue.
* For each of these, there's a "Shared" version (e.g. FullTypedSharedCacheInterface) that holds a shared_ptr to the Cache, rather than assuming external ownership by holding only a raw `Cache*`.

These interfaces introduce specific handle types for each interface instantiation, so that it's easy to see what kind of object is controlled by a handle. (Ultimately, this might not be worth the extra complexity, but it seems OK so far.)

Note: I attempted to make the cache 'charge' automatically inferred from the cache object type, such as by expecting an ApproximateMemoryUsage() function, but this is not so clean because there are cases where we need to compute the charge ahead of time and don't want to re-compute it.

## block_cache.h
This header is essentially the replacement for the old block_like_traits.h. It includes various things to support block cache access with typed_cache.h for block-based table.

## block_based_table_reader.cc
Before this change, accessing the block cache here was an awkward mix of static polymorphism (template TBlocklike) and switch-case on a dynamic BlockType value. This change mostly unifies on static polymorphism, relying on minor hacks in block_cache.h to distinguish variants of Block. We still check BlockType in some places (especially for stats, which could be improved in follow-up work) but at least the BlockType is a static constant from the template parameter. (No more awkward partial redundancy between static and dynamic info.) This likely contributes to the overall performance improvement, but hasn't been tested in isolation.

The other key source of simplification here is a more unified system of creating block cache objects: for directly populating from primary cache and for promotion from secondary cache. Both use BlockCreateContext, for context and for factory functions.

## block_based_table_builder.cc, cache_dump_load_impl.cc
Before this change, warming caches was super ugly code. Both of these source files had switch statements to basically transition from the dynamic BlockType world to the static TBlocklike world. None of that mess is needed anymore as there's a new, untyped WarmInCache function that handles all the details just as promotion from SecondaryCache would. (Fixes `TODO akanksha: Dedup below code` in block_based_table_builder.cc.)

## Everything else
Mostly just updating Cache users to use new typed APIs when reasonably possible, or changed Cache APIs when not.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/10975

Test Plan:
tests updated

Performance test setup similar to https://github.com/facebook/rocksdb/issues/10626 (by cache size, LRUCache when not "hyper" for HyperClockCache):

34MB 1thread base.hyper -> kops/s: 0.745 io_bytes/op: 2.52504e+06 miss_ratio: 0.140906 max_rss_mb: 76.4844
34MB 1thread new.hyper -> kops/s: 0.751 io_bytes/op: 2.5123e+06 miss_ratio: 0.140161 max_rss_mb: 79.3594
34MB 1thread base -> kops/s: 0.254 io_bytes/op: 1.36073e+07 miss_ratio: 0.918818 max_rss_mb: 45.9297
34MB 1thread new -> kops/s: 0.252 io_bytes/op: 1.36157e+07 miss_ratio: 0.918999 max_rss_mb: 44.1523
34MB 32thread base.hyper -> kops/s: 7.272 io_bytes/op: 2.88323e+06 miss_ratio: 0.162532 max_rss_mb: 516.602
34MB 32thread new.hyper -> kops/s: 7.214 io_bytes/op: 2.99046e+06 miss_ratio: 0.168818 max_rss_mb: 518.293
34MB 32thread base -> kops/s: 3.528 io_bytes/op: 1.35722e+07 miss_ratio: 0.914691 max_rss_mb: 264.926
34MB 32thread new -> kops/s: 3.604 io_bytes/op: 1.35744e+07 miss_ratio: 0.915054 max_rss_mb: 264.488
233MB 1thread base.hyper -> kops/s: 53.909 io_bytes/op: 2552.35 miss_ratio: 0.0440566 max_rss_mb: 241.984
233MB 1thread new.hyper -> kops/s: 62.792 io_bytes/op: 2549.79 miss_ratio: 0.044043 max_rss_mb: 241.922
233MB 1thread base -> kops/s: 1.197 io_bytes/op: 2.75173e+06 miss_ratio: 0.103093 max_rss_mb: 241.559
233MB 1thread new -> kops/s: 1.199 io_bytes/op: 2.73723e+06 miss_ratio: 0.10305 max_rss_mb: 240.93
233MB 32thread base.hyper -> kops/s: 1298.69 io_bytes/op: 2539.12 miss_ratio: 0.0440307 max_rss_mb: 371.418
233MB 32thread new.hyper -> kops/s: 1421.35 io_bytes/op: 2538.75 miss_ratio: 0.0440307 max_rss_mb: 347.273
233MB 32thread base -> kops/s: 9.693 io_bytes/op: 2.77304e+06 miss_ratio: 0.103745 max_rss_mb: 569.691
233MB 32thread new -> kops/s: 9.75 io_bytes/op: 2.77559e+06 miss_ratio: 0.103798 max_rss_mb: 552.82
1597MB 1thread base.hyper -> kops/s: 58.607 io_bytes/op: 1449.14 miss_ratio: 0.0249324 max_rss_mb: 1583.55
1597MB 1thread new.hyper -> kops/s: 69.6 io_bytes/op: 1434.89 miss_ratio: 0.0247167 max_rss_mb: 1584.02
1597MB 1thread base -> kops/s: 60.478 io_bytes/op: 1421.28 miss_ratio: 0.024452 max_rss_mb: 1589.45
1597MB 1thread new -> kops/s: 63.973 io_bytes/op: 1416.07 miss_ratio: 0.0243766 max_rss_mb: 1589.24
1597MB 32thread base.hyper -> kops/s: 1436.2 io_bytes/op: 1357.93 miss_ratio: 0.0235353 max_rss_mb: 1692.92
1597MB 32thread new.hyper -> kops/s: 1605.03 io_bytes/op: 1358.04 miss_ratio: 0.023538 max_rss_mb: 1702.78
1597MB 32thread base -> kops/s: 280.059 io_bytes/op: 1350.34 miss_ratio: 0.023289 max_rss_mb: 1675.36
1597MB 32thread new -> kops/s: 283.125 io_bytes/op: 1351.05 miss_ratio: 0.0232797 max_rss_mb: 1703.83

Almost uniformly improving over base revision, especially for hot paths with HyperClockCache, up to 12% higher throughput seen (1597MB, 32thread, hyper). The improvement for that is likely coming from much simplified code for providing context for secondary cache promotion (CreateCallback/CreateContext), and possibly from less branching in block_based_table_reader. And likely a small improvement from not reconstituting key for DeleterFn.

Reviewed By: anand1976

Differential Revision: D42417818

Pulled By: pdillinger

fbshipit-source-id: f86bfdd584dce27c028b151ba56818ad14f7a432
2023-01-11 14:20:40 -08:00

274 lines
13 KiB
C++

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
//
// Thread-safe (provides internal synchronization)
#pragma once
#include <cstdint>
#include <string>
#include <vector>
#include "cache/typed_cache.h"
#include "db/dbformat.h"
#include "db/range_del_aggregator.h"
#include "options/cf_options.h"
#include "port/port.h"
#include "rocksdb/cache.h"
#include "rocksdb/env.h"
#include "rocksdb/options.h"
#include "rocksdb/table.h"
#include "table/table_reader.h"
#include "trace_replay/block_cache_tracer.h"
#include "util/coro_utils.h"
namespace ROCKSDB_NAMESPACE {
class Env;
class Arena;
struct FileDescriptor;
class GetContext;
class HistogramImpl;
// Manages caching for TableReader objects for a column family. The actual
// cache is allocated separately and passed to the constructor. TableCache
// wraps around the underlying SST file readers by providing Get(),
// MultiGet() and NewIterator() methods that hide the instantiation,
// caching and access to the TableReader. The main purpose of this is
// performance - by caching the TableReader, it avoids unnecessary file opens
// and object allocation and instantiation. One exception is compaction, where
// a new TableReader may be instantiated - see NewIterator() comments
//
// Another service provided by TableCache is managing the row cache - if the
// DB is configured with a row cache, and the lookup key is present in the row
// cache, lookup is very fast. The row cache is obtained from
// ioptions.row_cache
class TableCache {
public:
TableCache(const ImmutableOptions& ioptions,
const FileOptions* storage_options, Cache* cache,
BlockCacheTracer* const block_cache_tracer,
const std::shared_ptr<IOTracer>& io_tracer,
const std::string& db_session_id);
~TableCache();
// Cache interface for table cache
using CacheInterface =
BasicTypedCacheInterface<TableReader, CacheEntryRole::kMisc>;
using TypedHandle = CacheInterface::TypedHandle;
// Cache interface for row cache
using RowCacheInterface =
BasicTypedCacheInterface<std::string, CacheEntryRole::kMisc>;
using RowHandle = RowCacheInterface::TypedHandle;
// Return an iterator for the specified file number (the corresponding
// file length must be exactly "file_size" bytes). If "table_reader_ptr"
// is non-nullptr, also sets "*table_reader_ptr" to point to the Table object
// underlying the returned iterator, or nullptr if no Table object underlies
// the returned iterator. The returned "*table_reader_ptr" object is owned
// by the cache and should not be deleted, and is valid for as long as the
// returned iterator is live.
// If !options.ignore_range_deletions, and range_del_iter is non-nullptr,
// then range_del_iter is set to a TruncatedRangeDelIterator for range
// tombstones in the SST file corresponding to the specified file number. The
// upper/lower bounds for the TruncatedRangeDelIterator are set to the SST
// file's boundary.
// @param options Must outlive the returned iterator.
// @param range_del_agg If non-nullptr, adds range deletions to the
// aggregator. If an error occurs, returns it in a NewErrorInternalIterator
// @param for_compaction If true, a new TableReader may be allocated (but
// not cached), depending on the CF options
// @param skip_filters Disables loading/accessing the filter block
// @param level The level this table is at, -1 for "not set / don't know"
InternalIterator* NewIterator(
const ReadOptions& options, const FileOptions& toptions,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta, RangeDelAggregator* range_del_agg,
const std::shared_ptr<const SliceTransform>& prefix_extractor,
TableReader** table_reader_ptr, HistogramImpl* file_read_hist,
TableReaderCaller caller, Arena* arena, bool skip_filters, int level,
size_t max_file_size_for_l0_meta_pin,
const InternalKey* smallest_compaction_key,
const InternalKey* largest_compaction_key, bool allow_unprepared_value,
TruncatedRangeDelIterator** range_del_iter = nullptr);
// If a seek to internal key "k" in specified file finds an entry,
// call get_context->SaveValue() repeatedly until
// it returns false. As a side effect, it will insert the TableReader
// into the cache and potentially evict another entry
// @param get_context Context for get operation. The result of the lookup
// can be retrieved by calling get_context->State()
// @param file_read_hist If non-nullptr, the file reader statistics are
// recorded
// @param skip_filters Disables loading/accessing the filter block
// @param level The level this table is at, -1 for "not set / don't know"
Status 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 = nullptr,
HistogramImpl* file_read_hist = nullptr, bool skip_filters = false,
int level = -1, size_t max_file_size_for_l0_meta_pin = 0);
// Return the range delete tombstone iterator of the file specified by
// `file_meta`.
Status GetRangeTombstoneIterator(
const ReadOptions& options,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta,
std::unique_ptr<FragmentedRangeTombstoneIterator>* out_iter);
// Call table reader's MultiGetFilter to use the bloom filter to filter out
// keys. Returns Status::NotSupported() if row cache needs to be checked.
// If the table cache is looked up to get the table reader, the cache handle
// is returned in table_handle. This handle should be passed back to
// MultiGet() so it can be released.
Status MultiGetFilter(
const ReadOptions& options,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta,
const std::shared_ptr<const SliceTransform>& prefix_extractor,
HistogramImpl* file_read_hist, int level,
MultiGetContext::Range* mget_range, TypedHandle** table_handle);
// If a seek to internal key "k" in specified file finds an entry,
// call get_context->SaveValue() repeatedly until
// it returns false. As a side effect, it will insert the TableReader
// into the cache and potentially evict another entry
// @param mget_range Pointer to the structure describing a batch of keys to
// be looked up in this table file. The result is stored
// in the embedded GetContext
// @param skip_filters Disables loading/accessing the filter block
// @param level The level this table is at, -1 for "not set / don't know"
DECLARE_SYNC_AND_ASYNC(
Status, MultiGet, const ReadOptions& options,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta, const MultiGetContext::Range* mget_range,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr,
HistogramImpl* file_read_hist = nullptr, bool skip_filters = false,
bool skip_range_deletions = false, int level = -1,
TypedHandle* table_handle = nullptr);
// Evict any entry for the specified file number
static void Evict(Cache* cache, uint64_t file_number);
// Find table reader
// @param skip_filters Disables loading/accessing the filter block
// @param level == -1 means not specified
Status FindTable(
const ReadOptions& ro, const FileOptions& toptions,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta, TypedHandle**,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr,
const bool no_io = false, bool record_read_stats = true,
HistogramImpl* file_read_hist = nullptr, bool skip_filters = false,
int level = -1, bool prefetch_index_and_filter_in_cache = true,
size_t max_file_size_for_l0_meta_pin = 0,
Temperature file_temperature = Temperature::kUnknown);
// Get the table properties of a given table.
// @no_io: indicates if we should load table to the cache if it is not present
// in table cache yet.
// @returns: `properties` will be reset on success. Please note that we will
// return Status::Incomplete() if table is not present in cache and
// we set `no_io` to be true.
Status GetTableProperties(
const FileOptions& toptions,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta,
std::shared_ptr<const TableProperties>* properties,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr,
bool no_io = false);
Status ApproximateKeyAnchors(const ReadOptions& ro,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta,
std::vector<TableReader::Anchor>& anchors);
// Return total memory usage of the table reader of the file.
// 0 if table reader of the file is not loaded.
size_t GetMemoryUsageByTableReader(
const FileOptions& toptions,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr);
// Returns approximated offset of a key in a file represented by fd.
uint64_t ApproximateOffsetOf(
const Slice& key, const FileMetaData& file_meta, TableReaderCaller caller,
const InternalKeyComparator& internal_comparator,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr);
// Returns approximated data size between start and end keys in a file
// represented by fd (the start key must not be greater than the end key).
uint64_t ApproximateSize(
const Slice& start, const Slice& end, const FileMetaData& file_meta,
TableReaderCaller caller,
const InternalKeyComparator& internal_comparator,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr);
CacheInterface& get_cache() { return cache_; }
// Capacity of the backing Cache that indicates infinite TableCache capacity.
// For example when max_open_files is -1 we set the backing Cache to this.
static const int kInfiniteCapacity = 0x400000;
// The tables opened with this TableCache will be immortal, i.e., their
// lifetime is as long as that of the DB.
void SetTablesAreImmortal() {
if (cache_.get()->GetCapacity() >= kInfiniteCapacity) {
immortal_tables_ = true;
}
}
private:
// Build a table reader
Status GetTableReader(
const ReadOptions& ro, const FileOptions& file_options,
const InternalKeyComparator& internal_comparator,
const FileMetaData& file_meta, bool sequential_mode,
bool record_read_stats, HistogramImpl* file_read_hist,
std::unique_ptr<TableReader>* table_reader,
const std::shared_ptr<const SliceTransform>& prefix_extractor = nullptr,
bool skip_filters = false, int level = -1,
bool prefetch_index_and_filter_in_cache = true,
size_t max_file_size_for_l0_meta_pin = 0,
Temperature file_temperature = Temperature::kUnknown);
// Update the max_covering_tombstone_seq in the GetContext for each key based
// on the range deletions in the table
void UpdateRangeTombstoneSeqnums(const ReadOptions& options, TableReader* t,
MultiGetContext::Range& table_range);
// Create a key prefix for looking up the row cache. The prefix is of the
// format row_cache_id + fd_number + seq_no. Later, the user key can be
// appended to form the full key
void CreateRowCacheKeyPrefix(const ReadOptions& options,
const FileDescriptor& fd,
const Slice& internal_key,
GetContext* get_context, IterKey& row_cache_key);
// Helper function to lookup the row cache for a key. It appends the
// user key to row_cache_key at offset prefix_size
bool GetFromRowCache(const Slice& user_key, IterKey& row_cache_key,
size_t prefix_size, GetContext* get_context);
const ImmutableOptions& ioptions_;
const FileOptions& file_options_;
CacheInterface cache_;
std::string row_cache_id_;
bool immortal_tables_;
BlockCacheTracer* const block_cache_tracer_;
Striped<port::Mutex, Slice> loader_mutex_;
std::shared_ptr<IOTracer> io_tracer_;
std::string db_session_id_;
};
} // namespace ROCKSDB_NAMESPACE