YugabyteDB (2.13.1.0-b60, 21121d69985fbf76aa6958d8f04a9bfa936293b5)

// Copyright (c) YugaByte, Inc.

//

// Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except

// in compliance with the License.  You may obtain a copy of the License at

//

// http://www.apache.org/licenses/LICENSE-2.0

//

// Unless required by applicable law or agreed to in writing, software distributed under the License

// is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express

// or implied.  See the License for the specific language governing permissions and limitations

// under the License.

//

#ifndef YB_DOCDB_DOCDB_ROCKSDB_UTIL_H_

#define YB_DOCDB_DOCDB_ROCKSDB_UTIL_H_

#include <boost/optional.hpp>

#include "yb/docdb/bounded_rocksdb_iterator.h"

#include "yb/rocksdb/cache.h"

#include "yb/rocksdb/db.h"

#include "yb/rocksdb/options.h"

#include "yb/rocksdb/rate_limiter.h"

#include "yb/rocksdb/table.h"

#include "yb/tablet/tablet_options.h"

#include "yb/util/slice.h"

namespace yb {

namespace docdb {

class IntentAwareIterator;

// See to a rocksdb point that is at least sub_doc_key.

// If the iterator is already positioned far enough, does not perform a seek.

void SeekForward(const rocksdb::Slice& slice, rocksdb::Iterator *iter);

void SeekForward(const KeyBytes& key_bytes, rocksdb::Iterator *iter);

// When we replace HybridTime::kMin in the end of seek key, next seek will skip older versions of

// this key, but will not skip any subkeys in its subtree. If the iterator is already positioned far

// enough, does not perform a seek.

void SeekPastSubKey(const Slice& key, rocksdb::Iterator* iter);

// Seek out of the given SubDocKey. For efficiency, the method that takes a non-const KeyBytes

// pointer avoids memory allocation by using the KeyBytes buffer to prepare the key to seek to by

// appending an extra byte. The appended byte is removed when the method returns.

void SeekOutOfSubKey(KeyBytes* key_bytes, rocksdb::Iterator* iter);

KeyBytes AppendDocHt(const Slice& key, const DocHybridTime& doc_ht);

// A wrapper around the RocksDB seek operation that uses Next() up to the configured number of

// times to avoid invalidating iterator state. In debug mode it also allows printing detailed

// information about RocksDB seeks.

void PerformRocksDBSeek(

    rocksdb::Iterator *iter,

    const rocksdb::Slice &seek_key,

    const char* file_name,

    int line);

// TODO: is there too much overhead in passing file name and line here in release mode?

#define ROCKSDB_SEEK(iter, key) \

  do { \

    PerformRocksDBSeek((iter), (key), __FILE__, __LINE__); \

  } while (0)

enum class BloomFilterMode {

  USE_BLOOM_FILTER,

  DONT_USE_BLOOM_FILTER,

};

// It is only allowed to use bloom filters on scans within the same hashed components of the key,

// because BloomFilterAwareIterator relies on it and ignores SST file completely if there are no

// keys with the same hashed components as key specified for seek operation.

// Note: bloom_filter_mode should be specified explicitly to avoid using it incorrectly by default.

// user_key_for_filter is used with BloomFilterMode::USE_BLOOM_FILTER to exclude SST files which

// have the same hashed components as (Sub)DocKey encoded in user_key_for_filter.

BoundedRocksDbIterator CreateRocksDBIterator(

    rocksdb::DB* rocksdb,

    const KeyBounds* docdb_key_bounds,

    BloomFilterMode bloom_filter_mode,

    const boost::optional<const Slice>& user_key_for_filter,

    const rocksdb::QueryId query_id,

    std::shared_ptr<rocksdb::ReadFileFilter> file_filter = nullptr,

    const Slice* iterate_upper_bound = nullptr);

// Values and transactions committed later than high_ht can be skipped, so we won't spend time

// for re-requesting pending transaction status if we already know it wasn't committed at high_ht.

std::unique_ptr<IntentAwareIterator> CreateIntentAwareIterator(

    const DocDB& doc_db,

    BloomFilterMode bloom_filter_mode,

    const boost::optional<const Slice>& user_key_for_filter,

    const rocksdb::QueryId query_id,

    const TransactionOperationContext& transaction_context,

    CoarseTimePoint deadline,

    const ReadHybridTime& read_time,

    std::shared_ptr<rocksdb::ReadFileFilter> file_filter = nullptr,

    const Slice* iterate_upper_bound = nullptr);

// Request RocksDB compaction and wait until it completes.

CHECKED_STATUS ForceRocksDBCompact(rocksdb::DB* db);

rocksdb::Options TEST_AutoInitFromRocksDBFlags();

rocksdb::BlockBasedTableOptions TEST_AutoInitFromRocksDbTableFlags();

Result<rocksdb::KeyValueEncodingFormat> GetConfiguredKeyValueEncodingFormat(

    const std::string& flag_value);

// Defines how rate limiter is shared across a node

YB_DEFINE_ENUM(RateLimiterSharingMode, (NONE)(TSERVER));

// Extracts rate limiter's sharing mode depending on the value of

// flag `FLAGS_rocksdb_compact_flush_rate_limit_sharing_mode`;

// `RateLimiterSharingMode::NONE` is returned if extraction failed

RateLimiterSharingMode GetRocksDBRateLimiterSharingMode();

// Creates `rocksdb::RateLimiter` taking into account related GFlags,

// calls `rocksdb::NewGenericRateLimiter` internally

std::shared_ptr<rocksdb::RateLimiter> CreateRocksDBRateLimiter();

// Initialize the RocksDB 'options'.

// The 'statistics' object provided by the caller will be used by RocksDB to maintain the stats for

// the tablet.

void InitRocksDBOptions(

    rocksdb::Options* options, const std::string& log_prefix,

    const std::shared_ptr<rocksdb::Statistics>& statistics,

    const tablet::TabletOptions& tablet_options,

    rocksdb::BlockBasedTableOptions table_options = rocksdb::BlockBasedTableOptions());

// Sets logs prefix for RocksDB options. This will also reinitialize options->info_log.

void SetLogPrefix(rocksdb::Options* options, const std::string& log_prefix);

// Gets the configured size of the node-global RocksDB priority thread pool.

int32_t GetGlobalRocksDBPriorityThreadPoolSize();

// Class to edit RocksDB manifest w/o fully loading DB into memory.

class RocksDBPatcher {

 public:

  explicit RocksDBPatcher(const std::string& dbpath, const rocksdb::Options& options);

  ~RocksDBPatcher();

  // Loads DB into patcher.

  CHECKED_STATUS Load();

  // Set hybrid time filter for DB.

  CHECKED_STATUS SetHybridTimeFilter(HybridTime value);

  // Modify flushed frontier and clean up smallest/largest op id in per-SST file metadata.

  CHECKED_STATUS ModifyFlushedFrontier(const ConsensusFrontier& frontier);

  // Update file sizes in manifest if actual file size was changed because of direct manipulation

  // with .sst files.

  // Like all other methods in this class it updates manifest file.

  CHECKED_STATUS UpdateFileSizes();

 private:

  class Impl;

  std::unique_ptr<Impl> impl_;

};

}  // namespace docdb

}  // namespace yb

#endif  // YB_DOCDB_DOCDB_ROCKSDB_UTIL_H_