Leveldb 元信息相关的数据结构解析1

LevelDB 元信息相关的数据结构解析1

0.如何实现的

0.1 Version

class Version {
 public:
  // Lookup the value for key.  If found, store it in *val and
  // return OK.  Else return a non-OK status.  Fills *stats.
  // REQUIRES: lock is not held
  struct GetStats {
    FileMetaData* seek_file;
    int seek_file_level;
  };

  // Append to *iters a sequence of iterators that will
  // yield the contents of this Version when merged together.
  // REQUIRES: This version has been saved (see VersionSet::SaveTo)
  void AddIterators(const ReadOptions&, std::vector<Iterator*>* iters);

  Status Get(const ReadOptions&, const LookupKey& key, std::string* val,
             GetStats* stats);

  // Adds "stats" into the current state.  Returns true if a new
  // compaction may need to be triggered, false otherwise.
  // REQUIRES: lock is held
  bool UpdateStats(const GetStats& stats);

  // Record a sample of bytes read at the specified internal key.
  // Samples are taken approximately once every config::kReadBytesPeriod
  // bytes.  Returns true if a new compaction may need to be triggered.
  // REQUIRES: lock is held
  bool RecordReadSample(Slice key);

  // Reference count management (so Versions do not disappear out from
  // under live iterators)
  void Ref();
  void Unref();

  void GetOverlappingInputs(
      int level,
      const InternalKey* begin,  // nullptr means before all keys
      const InternalKey* end,    // nullptr means after all keys
      std::vector<FileMetaData*>* inputs);

  // Returns true iff some file in the specified level overlaps
  // some part of [*smallest_user_key,*largest_user_key].
  // smallest_user_key==nullptr represents a key smaller than all the DB's keys.
  // largest_user_key==nullptr represents a key largest than all the DB's keys.
  bool OverlapInLevel(int level, const Slice* smallest_user_key,
                      const Slice* largest_user_key);

  // Return the level at which we should place a new memtable compaction
  // result that covers the range [smallest_user_key,largest_user_key].
  // 用来选择将从MemTable dump出的sstable file放入第几层
  int PickLevelForMemTableOutput(const Slice& smallest_user_key,
                                 const Slice& largest_user_key);

  // 判断某层level的文件个数
  int NumFiles(int level) const { return files_[level].size(); }

  // Return a human readable string that describes this version's contents.
  std::string DebugString() const;

 private:
  friend class Compaction;
  friend class VersionSet;

  class LevelFileNumIterator;

  explicit Version(VersionSet* vset)
      : vset_(vset),
        next_(this),
        prev_(this),
        refs_(0),
        file_to_compact_(nullptr),
        file_to_compact_level_(-1),
        compaction_score_(-1),
        compaction_level_(-1) {}

  Version(const Version&) = delete;
  Version& operator=(const Version&) = delete;

  ~Version();

  Iterator* NewConcatenatingIterator(const ReadOptions&, int level) const;

  // Call func(arg, level, f) for every file that overlaps user_key in
  // order from newest to oldest.  If an invocation of func returns
  // false, makes no more calls.
  //
  // REQUIRES: user portion of internal_key == user_key.
  void ForEachOverlapping(Slice user_key, Slice internal_key, void* arg,
                          bool (*func)(void*, int, FileMetaData*));

  VersionSet* vset_;  // VersionSet to which this Version belongs 所有的version都属于一个集合:VersionSet
  // VersionSet是Version组成的双链表,这里Version记录前一个和后一个Version两个指针
  Version* next_;     // Next version in linked list
  Version* prev_;     // Previous version in linked list
  int refs_;          // Number of live refs to this version

  // List of files per level
  // 这个二维数组,分层记录了所有的SST文件信息,每个SST文件由FileMetaData表示
  std::vector<FileMetaData*> files_[config::kNumLevels];

  // Next file to compact based on seek stats.
  FileMetaData* file_to_compact_;
  int file_to_compact_level_;

  // Level that should be compacted next and its compaction score.
  // Score < 1 means compaction is not strictly needed.  These fields
  // are initialized by Finalize().
  // Compaction需要用compaction_score_来判断是否需要发起major compaction,这部分逻辑与某level所有SSTable file的大小有关系
  double compaction_score_;
  int compaction_level_;
};
struct FileMetaData {
  FileMetaData() : refs(0), allowed_seeks(1 << 30), file_size(0) {}

  int refs;
  int allowed_seeks;  // Seeks allowed until compaction
  uint64_t number;
  uint64_t file_size;    // File size in bytes
  InternalKey smallest;  // Smallest internal key served by table
  InternalKey largest;   // Largest internal key served by table
};
Version::~Version() {
  assert(refs_ == 0);

  /* 从VersionSet中注销 */
  prev_->next_ = next_;
  next_->prev_ = prev_;

  /*本Version下所有的文件,引用计数减1*/
  for (int level = 0; level < config::kNumLevels; level++) {
    for (size_t i = 0; i < files_[level].size(); i++) {
      FileMetaData* f = files_[level][i];
      assert(f->refs > 0);
      f->refs--;
      if (f->refs <= 0) {
        delete f;
      }
    }
  }
}

0.2 VersionSet

class VersionSet {
 public:
  VersionSet(const std::string& dbname, const Options* options,
             TableCache* table_cache, const InternalKeyComparator*);
  VersionSet(const VersionSet&) = delete;
  VersionSet& operator=(const VersionSet&) = delete;

  ~VersionSet();

  // Apply *edit to the current version to form a new descriptor that
  // is both saved to persistent state and installed as the new
  // current version.  Will release *mu while actually writing to the file.
  // REQUIRES: *mu is held on entry.
  // REQUIRES: no other thread concurrently calls LogAndApply()
  Status LogAndApply(VersionEdit* edit, port::Mutex* mu)
      EXCLUSIVE_LOCKS_REQUIRED(mu);

  // Recover the last saved descriptor from persistent storage.
  Status Recover(bool* save_manifest);

  // Return the current version.
  Version* current() const { return current_; }

  // Return the current manifest file number
  uint64_t ManifestFileNumber() const { return manifest_file_number_; }

  // Allocate and return a new file number
  uint64_t NewFileNumber() { return next_file_number_++; }

  // Arrange to reuse "file_number" unless a newer file number has
  // already been allocated.
  // REQUIRES: "file_number" was returned by a call to NewFileNumber().
  void ReuseFileNumber(uint64_t file_number) {
    if (next_file_number_ == file_number + 1) {
      next_file_number_ = file_number;
    }
  }

  // Return the number of Table files at the specified level.
  int NumLevelFiles(int level) const;

  // Return the combined file size of all files at the specified level.
  int64_t NumLevelBytes(int level) const;

  // Return the last sequence number.
  uint64_t LastSequence() const { return last_sequence_; }

  // Set the last sequence number to s.
  void SetLastSequence(uint64_t s) {
    assert(s >= last_sequence_);
    last_sequence_ = s;
  }

  // Mark the specified file number as used.
  void MarkFileNumberUsed(uint64_t number);

  // Return the current log file number.
  uint64_t LogNumber() const { return log_number_; }

  // Return the log file number for the log file that is currently
  // being compacted, or zero if there is no such log file.
  uint64_t PrevLogNumber() const { return prev_log_number_; }

  // Pick level and inputs for a new compaction.
  // Returns nullptr if there is no compaction to be done.
  // Otherwise returns a pointer to a heap-allocated object that
  // describes the compaction.  Caller should delete the result.
  Compaction* PickCompaction();

  // Return a compaction object for compacting the range [begin,end] in
  // the specified level.  Returns nullptr if there is nothing in that
  // level that overlaps the specified range.  Caller should delete
  // the result.
  Compaction* CompactRange(int level, const InternalKey* begin,
                           const InternalKey* end);

  // Return the maximum overlapping data (in bytes) at next level for any
  // file at a level >= 1.
  int64_t MaxNextLevelOverlappingBytes();

  // Create an iterator that reads over the compaction inputs for "*c".
  // The caller should delete the iterator when no longer needed.
  Iterator* MakeInputIterator(Compaction* c);

  // Returns true iff some level needs a compaction.
  bool NeedsCompaction() const {
    Version* v = current_;
    return (v->compaction_score_ >= 1) || (v->file_to_compact_ != nullptr);
  }

  // Add all files listed in any live version to *live.
  // May also mutate some internal state.
  void AddLiveFiles(std::set<uint64_t>* live);

  // Return the approximate offset in the database of the data for
  // "key" as of version "v".
  uint64_t ApproximateOffsetOf(Version* v, const InternalKey& key);

  // Return a human-readable short (single-line) summary of the number
  // of files per level.  Uses *scratch as backing store.
  struct LevelSummaryStorage {
    char buffer[100];
  };
  const char* LevelSummary(LevelSummaryStorage* scratch) const;

 private:
  class Builder;

  friend class Compaction;
  friend class Version;

  bool ReuseManifest(const std::string& dscname, const std::string& dscbase);

  void Finalize(Version* v);

  void GetRange(const std::vector<FileMetaData*>& inputs, InternalKey* smallest,
                InternalKey* largest);

  void GetRange2(const std::vector<FileMetaData*>& inputs1,
                 const std::vector<FileMetaData*>& inputs2,
                 InternalKey* smallest, InternalKey* largest);

  void SetupOtherInputs(Compaction* c);

  // Save current contents to *log
  Status WriteSnapshot(log::Writer* log);

  void AppendVersion(Version* v);

  Env* const env_;
  const std::string dbname_;
  const Options* const options_;
  TableCache* const table_cache_;
  const InternalKeyComparator icmp_;
  uint64_t next_file_number_;
  uint64_t manifest_file_number_;
  uint64_t last_sequence_;
  uint64_t log_number_;
  uint64_t prev_log_number_;  // 0 or backing store for memtable being compacted

  // Opened lazily
  WritableFile* descriptor_file_;
  log::Writer* descriptor_log_;
  Version dummy_versions_;  // Head of circular doubly-linked list of versions.
  Version* current_;        // == dummy_versions_.prev_

  // Per-level key at which the next compaction at that level should start.
  // Either an empty string, or a valid InternalKey.
  std::string compact_pointer_[config::kNumLevels];
};

0.3VersionEdit

private: friend class VersionSet;

typedef std::set<std::pair<int, uint64_t» DeletedFileSet;

std::string comparator_; uint64t log_number; uint64t prev_log_number; uint64t next_file_number; SequenceNumber last_sequence_; bool has_comparator_; bool has_log_number_; bool has_prev_log_number_; bool has_next_file_number_; bool has_last_sequence_;

std::vector<std::pair<int, InternalKey» compact_pointers_; DeletedFileSet deleted_files_; std::vector<std::pair<int, FileMetaData» new_files_; }; ```