rocksdb/table/plain_table_index.h
Lei Jin 5665e5e285 introduce ImmutableOptions
Summary:
As a preparation to support updating some options dynamically, I'd like
to first introduce ImmutableOptions, which is a subset of Options that
cannot be changed during the course of a DB lifetime without restart.

ColumnFamily will keep both Options and ImmutableOptions. Any component
below ColumnFamily should only take ImmutableOptions in their
constructor. Other options should be taken from APIs, which will be
allowed to adjust dynamically.

I am yet to make changes to memtable and other related classes to take
ImmutableOptions in their ctor. That can be done in a seprate diff as
this one is already pretty big.

Test Plan: make all check

Reviewers: yhchiang, igor, sdong

Reviewed By: sdong

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D22545
2014-09-04 16:18:36 -07:00

222 lines
7.0 KiB
C++

// Copyright (c) 2014, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
#pragma once
#include <string>
#include <vector>
#include "db/dbformat.h"
#include "rocksdb/options.h"
#include "util/murmurhash.h"
#include "util/hash.h"
#include "util/arena.h"
#include "util/histogram.h"
namespace rocksdb {
// PlainTableIndex contains buckets size of index_size_, each is a
// 32-bit integer. The lower 31 bits contain an offset value (explained below)
// and the first bit of the integer indicates type of the offset.
//
// +--------------+------------------------------------------------------+
// | Flag (1 bit) | Offset to binary search buffer or file (31 bits) +
// +--------------+------------------------------------------------------+
//
// Explanation for the "flag bit":
//
// 0 indicates that the bucket contains only one prefix (no conflict when
// hashing this prefix), whose first row starts from this offset of the
// file.
// 1 indicates that the bucket contains more than one prefixes, or there
// are too many rows for one prefix so we need a binary search for it. In
// this case, the offset indicates the offset of sub_index_ holding the
// binary search indexes of keys for those rows. Those binary search indexes
// are organized in this way:
//
// The first 4 bytes, indicate how many indexes (N) are stored after it. After
// it, there are N 32-bit integers, each points of an offset of the file,
// which
// points to starting of a row. Those offsets need to be guaranteed to be in
// ascending order so the keys they are pointing to are also in ascending
// order
// to make sure we can use them to do binary searches. Below is visual
// presentation of a bucket.
//
// <begin>
// number_of_records: varint32
// record 1 file offset: fixedint32
// record 2 file offset: fixedint32
// ....
// record N file offset: fixedint32
// <end>
class PlainTableIndex {
public:
enum IndexSearchResult {
kNoPrefixForBucket = 0,
kDirectToFile = 1,
kSubindex = 2
};
explicit PlainTableIndex(Slice data) { InitFromRawData(data); }
PlainTableIndex()
: index_size_(0),
sub_index_size_(0),
num_prefixes_(0),
index_(nullptr),
sub_index_(nullptr) {}
IndexSearchResult GetOffset(uint32_t prefix_hash,
uint32_t* bucket_value) const;
Status InitFromRawData(Slice data);
const char* GetSubIndexBasePtrAndUpperBound(uint32_t offset,
uint32_t* upper_bound) const {
const char* index_ptr = &sub_index_[offset];
return GetVarint32Ptr(index_ptr, index_ptr + 4, upper_bound);
}
uint32_t GetIndexSize() const { return index_size_; }
uint32_t GetSubIndexSize() const { return sub_index_size_; }
uint32_t GetNumPrefixes() const { return num_prefixes_; }
static const uint64_t kMaxFileSize = (1u << 31) - 1;
static const uint32_t kSubIndexMask = 0x80000000;
static const size_t kOffsetLen = sizeof(uint32_t);
private:
uint32_t index_size_;
size_t sub_index_size_;
uint32_t num_prefixes_;
uint32_t* index_;
char* sub_index_;
};
// PlainTableIndexBuilder is used to create plain table index.
// After calling Finish(), it returns Slice, which is usually
// used either to initialize PlainTableIndex or
// to save index to sst file.
// For more details about the index, please refer to:
// https://github.com/facebook/rocksdb/wiki/PlainTable-Format
// #wiki-in-memory-index-format
class PlainTableIndexBuilder {
public:
PlainTableIndexBuilder(Arena* arena, const ImmutableCFOptions& ioptions,
uint32_t index_sparseness, double hash_table_ratio,
double huge_page_tlb_size)
: arena_(arena),
ioptions_(ioptions),
record_list_(kRecordsPerGroup),
is_first_record_(true),
due_index_(false),
num_prefixes_(0),
num_keys_per_prefix_(0),
prev_key_prefix_hash_(0),
index_sparseness_(index_sparseness),
prefix_extractor_(ioptions.prefix_extractor),
hash_table_ratio_(hash_table_ratio),
huge_page_tlb_size_(huge_page_tlb_size) {}
void AddKeyPrefix(Slice key_prefix_slice, uint64_t key_offset);
Slice Finish();
uint32_t GetTotalSize() const {
return VarintLength(index_size_) + VarintLength(num_prefixes_) +
PlainTableIndex::kOffsetLen * index_size_ + sub_index_size_;
}
static const std::string kPlainTableIndexBlock;
private:
struct IndexRecord {
uint32_t hash; // hash of the prefix
uint32_t offset; // offset of a row
IndexRecord* next;
};
// Helper class to track all the index records
class IndexRecordList {
public:
explicit IndexRecordList(size_t num_records_per_group)
: kNumRecordsPerGroup(num_records_per_group),
current_group_(nullptr),
num_records_in_current_group_(num_records_per_group) {}
~IndexRecordList() {
for (size_t i = 0; i < groups_.size(); i++) {
delete[] groups_[i];
}
}
void AddRecord(murmur_t hash, uint32_t offset);
size_t GetNumRecords() const {
return (groups_.size() - 1) * kNumRecordsPerGroup +
num_records_in_current_group_;
}
IndexRecord* At(size_t index) {
return &(groups_[index / kNumRecordsPerGroup]
[index % kNumRecordsPerGroup]);
}
private:
IndexRecord* AllocateNewGroup() {
IndexRecord* result = new IndexRecord[kNumRecordsPerGroup];
groups_.push_back(result);
return result;
}
// Each group in `groups_` contains fix-sized records (determined by
// kNumRecordsPerGroup). Which can help us minimize the cost if resizing
// occurs.
const size_t kNumRecordsPerGroup;
IndexRecord* current_group_;
// List of arrays allocated
std::vector<IndexRecord*> groups_;
size_t num_records_in_current_group_;
};
void AllocateIndex();
// Internal helper function to bucket index record list to hash buckets.
void BucketizeIndexes(std::vector<IndexRecord*>* hash_to_offsets,
std::vector<uint32_t>* entries_per_bucket);
// Internal helper class to fill the indexes and bloom filters to internal
// data structures.
Slice FillIndexes(const std::vector<IndexRecord*>& hash_to_offsets,
const std::vector<uint32_t>& entries_per_bucket);
Arena* arena_;
const ImmutableCFOptions ioptions_;
HistogramImpl keys_per_prefix_hist_;
IndexRecordList record_list_;
bool is_first_record_;
bool due_index_;
uint32_t num_prefixes_;
uint32_t num_keys_per_prefix_;
uint32_t prev_key_prefix_hash_;
uint32_t index_sparseness_;
uint32_t index_size_;
size_t sub_index_size_;
const SliceTransform* prefix_extractor_;
double hash_table_ratio_;
double huge_page_tlb_size_;
std::string prev_key_prefix_;
static const size_t kRecordsPerGroup = 256;
};
}; // namespace rocksdb