rocksdb/table/block_based_table_factory.h
Yi Wu b32d087dbb Move MemoryAllocator option from Cache to BlockBasedTableOptions (#4676)
Summary:
Per offline discussion with siying, `MemoryAllocator` and `Cache` should be decouple. The idea is that memory allocator handles memory allocation, while cache handle cache policy.

It is normal that external cache libraries pack couple the two components for better optimization. If we want to integrate with such library in the future, we can make a wrapper of the library implementing both `Cache` and `MemoryAllocator` interface.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4676

Differential Revision: D13047662

Pulled By: yiwu-arbug

fbshipit-source-id: cd42e246d80ab600b4de47d073f7d2db308ce6dd
2018-11-13 13:48:38 -08:00

194 lines
8.1 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.
#pragma once
#include <stdint.h>
#include <memory>
#include <string>
#include "db/dbformat.h"
#include "options/options_helper.h"
#include "options/options_parser.h"
#include "rocksdb/flush_block_policy.h"
#include "rocksdb/table.h"
namespace rocksdb {
struct EnvOptions;
using std::unique_ptr;
class BlockBasedTableBuilder;
// A class used to track actual bytes written from the tail in the recent SST
// file opens, and provide a suggestion for following open.
class TailPrefetchStats {
public:
void RecordEffectiveSize(size_t len);
// 0 indicates no information to determine.
size_t GetSuggestedPrefetchSize();
private:
const static size_t kNumTracked = 32;
size_t records_[kNumTracked];
port::Mutex mutex_;
size_t next_ = 0;
size_t num_records_ = 0;
};
class BlockBasedTableFactory : public TableFactory {
public:
explicit BlockBasedTableFactory(
const BlockBasedTableOptions& table_options = BlockBasedTableOptions());
~BlockBasedTableFactory() {}
const char* Name() const override { return kName.c_str(); }
Status NewTableReader(
const TableReaderOptions& table_reader_options,
std::unique_ptr<RandomAccessFileReader>&& file, uint64_t file_size,
std::unique_ptr<TableReader>* table_reader,
bool prefetch_index_and_filter_in_cache = true) const override;
TableBuilder* NewTableBuilder(
const TableBuilderOptions& table_builder_options,
uint32_t column_family_id, WritableFileWriter* file) const override;
// Sanitizes the specified DB Options.
Status SanitizeOptions(const DBOptions& db_opts,
const ColumnFamilyOptions& cf_opts) const override;
std::string GetPrintableTableOptions() const override;
Status GetOptionString(std::string* opt_string,
const std::string& delimiter) const override;
const BlockBasedTableOptions& table_options() const;
void* GetOptions() override { return &table_options_; }
bool IsDeleteRangeSupported() const override { return true; }
static const std::string kName;
private:
BlockBasedTableOptions table_options_;
mutable TailPrefetchStats tail_prefetch_stats_;
};
extern const std::string kHashIndexPrefixesBlock;
extern const std::string kHashIndexPrefixesMetadataBlock;
extern const std::string kPropTrue;
extern const std::string kPropFalse;
#ifndef ROCKSDB_LITE
extern Status VerifyBlockBasedTableFactory(
const BlockBasedTableFactory* base_tf,
const BlockBasedTableFactory* file_tf,
OptionsSanityCheckLevel sanity_check_level);
static std::unordered_map<std::string, OptionTypeInfo>
block_based_table_type_info = {
/* currently not supported
std::shared_ptr<Cache> block_cache = nullptr;
std::shared_ptr<Cache> block_cache_compressed = nullptr;
std::shared_ptr<MemoryAllocator> memory_allocator = nullptr;
*/
{"flush_block_policy_factory",
{offsetof(struct BlockBasedTableOptions, flush_block_policy_factory),
OptionType::kFlushBlockPolicyFactory, OptionVerificationType::kByName,
false, 0}},
{"cache_index_and_filter_blocks",
{offsetof(struct BlockBasedTableOptions,
cache_index_and_filter_blocks),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"cache_index_and_filter_blocks_with_high_priority",
{offsetof(struct BlockBasedTableOptions,
cache_index_and_filter_blocks_with_high_priority),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"pin_l0_filter_and_index_blocks_in_cache",
{offsetof(struct BlockBasedTableOptions,
pin_l0_filter_and_index_blocks_in_cache),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"index_type",
{offsetof(struct BlockBasedTableOptions, index_type),
OptionType::kBlockBasedTableIndexType,
OptionVerificationType::kNormal, false, 0}},
{"hash_index_allow_collision",
{offsetof(struct BlockBasedTableOptions, hash_index_allow_collision),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"data_block_index_type",
{offsetof(struct BlockBasedTableOptions, data_block_index_type),
OptionType::kBlockBasedTableDataBlockIndexType,
OptionVerificationType::kNormal, false, 0}},
{"data_block_hash_table_util_ratio",
{offsetof(struct BlockBasedTableOptions,
data_block_hash_table_util_ratio),
OptionType::kDouble, OptionVerificationType::kNormal, false, 0}},
{"checksum",
{offsetof(struct BlockBasedTableOptions, checksum),
OptionType::kChecksumType, OptionVerificationType::kNormal, false,
0}},
{"no_block_cache",
{offsetof(struct BlockBasedTableOptions, no_block_cache),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"block_size",
{offsetof(struct BlockBasedTableOptions, block_size),
OptionType::kSizeT, OptionVerificationType::kNormal, false, 0}},
{"block_size_deviation",
{offsetof(struct BlockBasedTableOptions, block_size_deviation),
OptionType::kInt, OptionVerificationType::kNormal, false, 0}},
{"block_restart_interval",
{offsetof(struct BlockBasedTableOptions, block_restart_interval),
OptionType::kInt, OptionVerificationType::kNormal, false, 0}},
{"index_block_restart_interval",
{offsetof(struct BlockBasedTableOptions, index_block_restart_interval),
OptionType::kInt, OptionVerificationType::kNormal, false, 0}},
{"index_per_partition",
{0, OptionType::kUInt64T, OptionVerificationType::kDeprecated, false,
0}},
{"metadata_block_size",
{offsetof(struct BlockBasedTableOptions, metadata_block_size),
OptionType::kUInt64T, OptionVerificationType::kNormal, false, 0}},
{"partition_filters",
{offsetof(struct BlockBasedTableOptions, partition_filters),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"filter_policy",
{offsetof(struct BlockBasedTableOptions, filter_policy),
OptionType::kFilterPolicy, OptionVerificationType::kByName, false,
0}},
{"whole_key_filtering",
{offsetof(struct BlockBasedTableOptions, whole_key_filtering),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"skip_table_builder_flush",
{0, OptionType::kBoolean, OptionVerificationType::kDeprecated, false,
0}},
{"format_version",
{offsetof(struct BlockBasedTableOptions, format_version),
OptionType::kUInt32T, OptionVerificationType::kNormal, false, 0}},
{"verify_compression",
{offsetof(struct BlockBasedTableOptions, verify_compression),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"read_amp_bytes_per_bit",
{offsetof(struct BlockBasedTableOptions, read_amp_bytes_per_bit),
OptionType::kSizeT, OptionVerificationType::kNormal, false, 0}},
{"enable_index_compression",
{offsetof(struct BlockBasedTableOptions, enable_index_compression),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"block_align",
{offsetof(struct BlockBasedTableOptions, block_align),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}},
{"pin_top_level_index_and_filter",
{offsetof(struct BlockBasedTableOptions,
pin_top_level_index_and_filter),
OptionType::kBoolean, OptionVerificationType::kNormal, false, 0}}};
#endif // !ROCKSDB_LITE
} // namespace rocksdb