2018-07-24 20:42:19 +02:00
|
|
|
// 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).
|
|
|
|
|
|
|
|
#pragma once
|
|
|
|
|
|
|
|
#include <string>
|
|
|
|
#include <vector>
|
|
|
|
|
2018-08-15 23:27:47 +02:00
|
|
|
#include "rocksdb/slice.h"
|
|
|
|
|
2018-07-24 20:42:19 +02:00
|
|
|
namespace rocksdb {
|
|
|
|
// This is an experimental feature aiming to reduce the CPU utilization of
|
2018-08-15 23:27:47 +02:00
|
|
|
// point-lookup within a data-block. It is only used in data blocks, and not
|
|
|
|
// in meta-data blocks or per-table index blocks.
|
|
|
|
//
|
|
|
|
// It only used to support BlockBasedTable::Get().
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
|
|
|
// A serialized hash index is appended to the data-block. The new block data
|
|
|
|
// format is as follows:
|
|
|
|
//
|
|
|
|
// DATA_BLOCK: [RI RI RI ... RI RI_IDX HASH_IDX FOOTER]
|
|
|
|
//
|
|
|
|
// RI: Restart Interval (the same as the default data-block format)
|
|
|
|
// RI_IDX: Restart Interval index (the same as the default data-block format)
|
|
|
|
// HASH_IDX: The new data-block hash index feature.
|
|
|
|
// FOOTER: A 32bit block footer, which is the NUM_RESTARTS with the MSB as
|
|
|
|
// the flag indicating if this hash index is in use. Note that
|
|
|
|
// given a data block < 32KB, the MSB is never used. So we can
|
2018-08-15 23:27:47 +02:00
|
|
|
// borrow the MSB as the hash index flag. Therefore, this format is
|
|
|
|
// compatible with the legacy data-blocks with num_restarts < 32768,
|
|
|
|
// as the MSB is 0.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// The format of the data-block hash index is as follows:
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// HASH_IDX: [B B B ... B NUM_BUCK]
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// B: bucket, an array of restart index. Each buckets is uint8_t.
|
|
|
|
// NUM_BUCK: Number of buckets, which is the length of the bucket array.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// We reserve two special flag:
|
|
|
|
// kNoEntry=255,
|
|
|
|
// kCollision=254.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// Therefore, the max number of restarts this hash index can supoport is 253.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// Buckets are initialized to be kNoEntry.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// When storing a key in the hash index, the key is first hashed to a bucket.
|
|
|
|
// If there the bucket is empty (kNoEntry), the restart index is stored in
|
|
|
|
// the bucket. If there is already a restart index there, we will update the
|
|
|
|
// existing restart index to a collision marker (kCollision). If the
|
|
|
|
// the bucket is already marked as collision, we do not store the restart
|
|
|
|
// index either.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// During query process, a key is first hashed to a bucket. Then we examine if
|
|
|
|
// the buckets store nothing (kNoEntry) or the bucket had a collision
|
|
|
|
// (kCollision). If either of those happens, we get the restart index of
|
|
|
|
// the key and will directly go to the restart interval to search the key.
|
2018-07-24 20:42:19 +02:00
|
|
|
//
|
2018-08-15 23:27:47 +02:00
|
|
|
// Note that we only support blocks with #restart_interval < 254. If a block
|
|
|
|
// has more restart interval than that, hash index will not be create for it.
|
|
|
|
|
|
|
|
const uint8_t kNoEntry = 255;
|
|
|
|
const uint8_t kCollision = 254;
|
|
|
|
const uint8_t kMaxRestartSupportedByHashIndex = 253;
|
|
|
|
|
|
|
|
// Because we use uint16_t address, we only support block no more than 64KB
|
|
|
|
const size_t kMaxBlockSizeSupportedByHashIndex = 1u << 16;
|
|
|
|
const double kDefaultUtilRatio = 0.75;
|
2018-07-24 20:42:19 +02:00
|
|
|
|
|
|
|
class DataBlockHashIndexBuilder {
|
|
|
|
public:
|
2018-08-21 08:04:08 +02:00
|
|
|
DataBlockHashIndexBuilder()
|
|
|
|
: bucket_per_key_(-1 /*uninitialized marker*/),
|
|
|
|
estimated_num_buckets_(0),
|
|
|
|
valid_(false) {}
|
2018-08-15 23:27:47 +02:00
|
|
|
|
|
|
|
void Initialize(double util_ratio) {
|
|
|
|
if (util_ratio <= 0) {
|
|
|
|
util_ratio = kDefaultUtilRatio; // sanity check
|
|
|
|
}
|
2018-08-21 08:04:08 +02:00
|
|
|
bucket_per_key_ = 1 / util_ratio;
|
2018-08-15 23:27:47 +02:00
|
|
|
valid_ = true;
|
|
|
|
}
|
|
|
|
|
2018-08-21 08:04:08 +02:00
|
|
|
inline bool Valid() const { return valid_ && bucket_per_key_ > 0; }
|
2018-08-15 23:27:47 +02:00
|
|
|
void Add(const Slice& key, const size_t restart_index);
|
2018-07-24 20:42:19 +02:00
|
|
|
void Finish(std::string& buffer);
|
|
|
|
void Reset();
|
2018-08-15 23:27:47 +02:00
|
|
|
inline size_t EstimateSize() const {
|
2018-08-21 08:04:08 +02:00
|
|
|
uint16_t estimated_num_buckets =
|
|
|
|
static_cast<uint16_t>(estimated_num_buckets_);
|
2018-08-15 23:27:47 +02:00
|
|
|
|
|
|
|
// Maching the num_buckets number in DataBlockHashIndexBuilder::Finish.
|
|
|
|
estimated_num_buckets |= 1;
|
|
|
|
|
|
|
|
return sizeof(uint16_t) +
|
|
|
|
static_cast<size_t>(estimated_num_buckets * sizeof(uint8_t));
|
|
|
|
}
|
2018-07-24 20:42:19 +02:00
|
|
|
|
|
|
|
private:
|
2018-08-21 08:04:08 +02:00
|
|
|
double bucket_per_key_; // is the multiplicative inverse of util_ratio_
|
|
|
|
double estimated_num_buckets_;
|
2018-08-15 23:27:47 +02:00
|
|
|
|
|
|
|
// Now the only usage for `valid_` is to mark false when the inserted
|
|
|
|
// restart_index is larger than supported. In this case HashIndex is not
|
|
|
|
// appended to the block content.
|
|
|
|
bool valid_;
|
2018-07-24 20:42:19 +02:00
|
|
|
|
2018-08-15 23:27:47 +02:00
|
|
|
std::vector<std::pair<uint32_t, uint8_t>> hash_and_restart_pairs_;
|
|
|
|
friend class DataBlockHashIndex_DataBlockHashTestSmall_Test;
|
|
|
|
};
|
2018-07-24 20:42:19 +02:00
|
|
|
|
|
|
|
class DataBlockHashIndex {
|
|
|
|
public:
|
2018-08-15 23:27:47 +02:00
|
|
|
DataBlockHashIndex() : num_buckets_(0) {}
|
2018-07-24 20:42:19 +02:00
|
|
|
|
2018-08-15 23:27:47 +02:00
|
|
|
void Initialize(const char* data, uint16_t size, uint16_t* map_offset);
|
|
|
|
|
|
|
|
uint8_t Lookup(const char* data, uint32_t map_offset, const Slice& key) const;
|
2018-07-24 20:42:19 +02:00
|
|
|
|
2018-08-15 23:27:47 +02:00
|
|
|
inline bool Valid() { return num_buckets_ != 0; }
|
2018-07-24 20:42:19 +02:00
|
|
|
|
|
|
|
private:
|
|
|
|
// To make the serialized hash index compact and to save the space overhead,
|
|
|
|
// here all the data fields persisted in the block are in uint16 format.
|
|
|
|
// We find that a uint16 is large enough to index every offset of a 64KiB
|
|
|
|
// block.
|
|
|
|
// So in other words, DataBlockHashIndex does not support block size equal
|
|
|
|
// or greater then 64KiB.
|
|
|
|
uint16_t num_buckets_;
|
|
|
|
};
|
|
|
|
|
|
|
|
} // namespace rocksdb
|