Revision 239d17a19c3cec16937aa4b6c56c90f4f217addf authored by Peter Dillinger on 18 December 2020, 22:29:48 UTC, committed by Facebook GitHub Bot on 18 December 2020, 22:31:03 UTC
Summary:
Primarily this change refactors the optimize_filters_for_memory
code for Bloom filters, based on malloc_usable_size, to also work for
Ribbon filters.

This change also replaces the somewhat slow but general
BuiltinFilterBitsBuilder::ApproximateNumEntries with
implementation-specific versions for Ribbon (new) and Legacy Bloom
(based on a recently deleted version). The reason is to emphasize
speed in ApproximateNumEntries rather than 100% accuracy.

Justification: ApproximateNumEntries (formerly CalculateNumEntry) is
only used by RocksDB for range-partitioned filters, called each time we
start to construct one. (In theory, it should be possible to reuse the
estimate, but the abstractions provided by FilterPolicy don't really
make that workable.) But this is only used as a heuristic estimate for
hitting a desired partitioned filter size because of alignment to data
blocks, which have various numbers of unique keys or prefixes. The two
factors lead us to prioritize reasonable speed over 100% accuracy.

optimize_filters_for_memory adds extra complication, because precisely
calculating num_entries for some allowed number of bytes depends on state
with optimize_filters_for_memory enabled. And the allocator-agnostic
implementation of optimize_filters_for_memory, using malloc_usable_size,
means we would have to actually allocate memory, many times, just to
precisely determine how many entries (keys) could be added and stay below
some size budget, for the current state. (In a draft, I got this
working, and then realized the balance of speed vs. accuracy was all
wrong.)

So related to that, I have made CalculateSpace, an internal-only API
only used for testing, non-authoritative also if
optimize_filters_for_memory is enabled. This simplifies some code.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7774

Test Plan:
unit test updated, and for FilterSize test, range of tested
values is greatly expanded (still super fast)

Also tested `db_bench -benchmarks=fillrandom,stats -bloom_bits=10 -num=1000000 -partition_index_and_filters -format_version=5 [-optimize_filters_for_memory] [-use_ribbon_filter]` with temporary debug output of generated filter sizes.

Bloom+optimize_filters_for_memory:

      1 Filter size: 197 (224 in memory)
    134 Filter size: 3525 (3584 in memory)
    107 Filter size: 4037 (4096 in memory)
    Total on disk: 904,506
    Total in memory: 918,752

Ribbon+optimize_filters_for_memory:

      1 Filter size: 3061 (3072 in memory)
    110 Filter size: 3573 (3584 in memory)
     58 Filter size: 4085 (4096 in memory)
    Total on disk: 633,021 (-30.0%)
    Total in memory: 634,880 (-30.9%)

Bloom (no offm):

      1 Filter size: 261 (320 in memory)
      1 Filter size: 3333 (3584 in memory)
    240 Filter size: 3717 (4096 in memory)
    Total on disk: 895,674 (-1% on disk vs. +offm; known tolerable overhead of offm)
    Total in memory: 986,944 (+7.4% vs. +offm)

Ribbon (no offm):

      1 Filter size: 2949 (3072 in memory)
      1 Filter size: 3381 (3584 in memory)
    167 Filter size: 3701 (4096 in memory)
    Total on disk: 624,397 (-30.3% vs. Bloom)
    Total in memory: 690,688 (-30.0% vs. Bloom)

Note that optimize_filters_for_memory is even more effective for Ribbon filter than for cache-local Bloom, because it can close the unused memory gap even tighter than Bloom filter, because of 16 byte increments for Ribbon vs. 64 byte increments for Bloom.

Reviewed By: jay-zhuang

Differential Revision: D25592970

Pulled By: pdillinger

fbshipit-source-id: 606fdaa025bb790d7e9c21601e8ea86e10541912
1 parent 04b3524
Raw File
block_fetcher.h
//  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.

#pragma once
#include "memory/memory_allocator.h"
#include "table/block_based/block.h"
#include "table/block_based/block_type.h"
#include "table/format.h"

namespace ROCKSDB_NAMESPACE {

// Retrieves a single block of a given file. Utilizes the prefetch buffer and/or
// persistent cache provided (if any) to try to avoid reading from the file
// directly. Note that both the prefetch buffer and the persistent cache are
// optional; also, note that the persistent cache may be configured to store either
// compressed or uncompressed blocks.
//
// If the retrieved block is compressed and the do_uncompress flag is set,
// BlockFetcher uncompresses the block (using the uncompression dictionary,
// if provided, to prime the compression algorithm), and returns the resulting
// uncompressed block data. Otherwise, it returns the original block.
//
// Two read options affect the behavior of BlockFetcher: if verify_checksums is
// true, the checksum of the (original) block is checked; if fill_cache is true,
// the block is added to the persistent cache if needed.
//
// Memory for uncompressed and compressed blocks is allocated as needed
// using memory_allocator and memory_allocator_compressed, respectively
// (if provided; otherwise, the default allocator is used).

class BlockFetcher {
 public:
  BlockFetcher(RandomAccessFileReader* file,
               FilePrefetchBuffer* prefetch_buffer, const Footer& footer,
               const ReadOptions& read_options, const BlockHandle& handle,
               BlockContents* contents, const ImmutableCFOptions& ioptions,
               bool do_uncompress, bool maybe_compressed, BlockType block_type,
               const UncompressionDict& uncompression_dict,
               const PersistentCacheOptions& cache_options,
               MemoryAllocator* memory_allocator = nullptr,
               MemoryAllocator* memory_allocator_compressed = nullptr,
               bool for_compaction = false)
      : file_(file),
        prefetch_buffer_(prefetch_buffer),
        footer_(footer),
        read_options_(read_options),
        handle_(handle),
        contents_(contents),
        ioptions_(ioptions),
        do_uncompress_(do_uncompress),
        maybe_compressed_(maybe_compressed),
        block_type_(block_type),
        block_size_(static_cast<size_t>(handle_.size())),
        block_size_with_trailer_(block_size(handle_)),
        uncompression_dict_(uncompression_dict),
        cache_options_(cache_options),
        memory_allocator_(memory_allocator),
        memory_allocator_compressed_(memory_allocator_compressed),
        for_compaction_(for_compaction) {}

  Status ReadBlockContents();
  CompressionType get_compression_type() const { return compression_type_; }

#ifndef NDEBUG
  int TEST_GetNumStackBufMemcpy() const { return num_stack_buf_memcpy_; }
  int TEST_GetNumHeapBufMemcpy() const { return num_heap_buf_memcpy_; }
  int TEST_GetNumCompressedBufMemcpy() const {
    return num_compressed_buf_memcpy_;
  }

#endif
 private:
#ifndef NDEBUG
  int num_stack_buf_memcpy_ = 0;
  int num_heap_buf_memcpy_ = 0;
  int num_compressed_buf_memcpy_ = 0;

#endif
  static const uint32_t kDefaultStackBufferSize = 5000;

  RandomAccessFileReader* file_;
  FilePrefetchBuffer* prefetch_buffer_;
  const Footer& footer_;
  const ReadOptions read_options_;
  const BlockHandle& handle_;
  BlockContents* contents_;
  const ImmutableCFOptions& ioptions_;
  const bool do_uncompress_;
  const bool maybe_compressed_;
  const BlockType block_type_;
  const size_t block_size_;
  const size_t block_size_with_trailer_;
  const UncompressionDict& uncompression_dict_;
  const PersistentCacheOptions& cache_options_;
  MemoryAllocator* memory_allocator_;
  MemoryAllocator* memory_allocator_compressed_;
  Status status_;
  Slice slice_;
  char* used_buf_ = nullptr;
  AlignedBuf direct_io_buf_;
  CacheAllocationPtr heap_buf_;
  CacheAllocationPtr compressed_buf_;
  char stack_buf_[kDefaultStackBufferSize];
  bool got_from_prefetch_buffer_ = false;
  CompressionType compression_type_;
  bool for_compaction_ = false;

  // return true if found
  bool TryGetUncompressBlockFromPersistentCache();
  // return true if found
  bool TryGetFromPrefetchBuffer();
  bool TryGetCompressedBlockFromPersistentCache();
  void PrepareBufferForBlockFromFile();
  // Copy content from used_buf_ to new heap_buf_.
  void CopyBufferToHeapBuf();
  // Copy content from used_buf_ to new compressed_buf_.
  void CopyBufferToCompressedBuf();
  void GetBlockContents();
  void InsertCompressedBlockToPersistentCacheIfNeeded();
  void InsertUncompressedBlockToPersistentCacheIfNeeded();
  void CheckBlockChecksum();
};
}  // namespace ROCKSDB_NAMESPACE
back to top