MySQL · RocksDB · 写入逻辑的实现

on 2018-07-04 | by 数据库内核月报关注作者

简介

在RocksDB中，每次写入它都会先写WAL,然后再写入MemTable,这次我们就来分析这两个逻辑具体是如何实现的. 首先需要明确的是在RocksDB中，WAL的写入是单线程顺序串行写入的，而MemTable则是可以并发多线程写入的。

而在RocksDB 5.5中引进了一个选项enable_pipelined_write，这个选项的目的就是将WAL和MemTable的写入pipeline化，也就是说当一个线程写完毕WAL之后，此时在WAL的write队列中等待的其他的write则会开始继续写入WAL, 而当前线程将会继续写入MemTable.此时就将不同的Writer的写入WAL和写入MemTable并发执行了.

实现

我们这里只来分析pipeline的实现，核心函数就是DBImpl::PipelinedWriteImpl.

每一个DB(DBImpl)都有一个write_thread_(class WriteThread). 每次调用Write的时候会先写入WAL, 此时新建一个WriteThread::Writer对象，并将这个对象加入到一个Group中(调用JoinBatchGroup)

WriteThread::Writer w(write_options, my_batch, callback, log_ref,
                      disable_memtable);
write_thread_.JoinBatchGroup(&amp;w);

然后我们来看JoinBatchGroup,这个函数主要是用来讲所有的写入WAL加入到一个Group中.这里可以看到当当前的Writer 对象是leader(比如第一个进入的对象)的时候将会直接返回，否则将会等待知道更新为对应的状态．

void WriteThread::JoinBatchGroup(Writer* w) {
...................................
bool linked_as_leader = LinkOne(w, &amp;newest_writer_);
if (linked_as_leader) {
  SetState(w, STATE_GROUP_LEADER);
}

TEST_SYNC_POINT_CALLBACK("WriteThread::JoinBatchGroup:Wait", w);

if (!linked_as_leader) {
  /**
   * Wait util:
   * 1) An existing leader pick us as the new leader when it finishes
   * 2) An existing leader pick us as its follewer and
   * 2.1) finishes the memtable writes on our behalf
   * 2.2) Or tell us to finish the memtable writes in pralallel
   * 3) (pipelined write) An existing leader pick us as its follower and
   *    finish book-keeping and WAL write for us, enqueue us as pending
   *    memtable writer, and
   * 3.1) we become memtable writer group leader, or
   * 3.2) an existing memtable writer group leader tell us to finish memtable
   *      writes in parallel.
   */
  AwaitState(w, STATE_GROUP_LEADER | STATE_MEMTABLE_WRITER_LEADER |
                    STATE_PARALLEL_MEMTABLE_WRITER | STATE_COMPLETED,
             &amp;jbg_ctx);
  TEST_SYNC_POINT_CALLBACK("WriteThread::JoinBatchGroup:DoneWaiting", w);
}
}

然后我们来看LinkOne函数，这个函数主要用来讲当前的Writer对象加入到group中，这里可以看到由于写入是并发的因此对应的newest_writer_(保存最新的写入对象)需要原子操作来更新.

bool WriteThread::LinkOne(Writer* w, std::atomic&lt;Writer*&gt;* newest_writer) {
assert(newest_writer != nullptr);
assert(w-&gt;state == STATE_INIT);
Writer* writers = newest_writer-&gt;load(std::memory_order_relaxed);
while (true) {
  w-&gt;link_older = writers;
  if (newest_writer-&gt;compare_exchange_weak(writers, w)) {
    return (writers == nullptr);
  }
}
}

当从JoinBatchGroup返回之后，当当前的Writer对象为leader的话，则将会把此leader下的所有的write都链接到一个WriteGroup中(调用EnterAsBatchGroupLeader函数),　并开始写入WAL,这里要注意非leader的write将会直接进入memtable的写入，这是因为非leader的write都将会被当前它所从属的leader来打包(group)写入，后面我们会看到实现.

size_t WriteThread::EnterAsBatchGroupLeader(Writer* leader,
                                          WriteGroup* write_group) {
assert(leader-&gt;link_older == nullptr);
assert(leader-&gt;batch != nullptr);
assert(write_group != nullptr);
................................................
Writer* newest_writer = newest_writer_.load(std::memory_order_acquire);

// This is safe regardless of any db mutex status of the caller. Previous
// calls to ExitAsGroupLeader either didn't call CreateMissingNewerLinks
// (they emptied the list and then we added ourself as leader) or had to
// explicitly wake us up (the list was non-empty when we added ourself,
// so we have already received our MarkJoined).
CreateMissingNewerLinks(newest_writer);

// Tricky. Iteration start (leader) is exclusive and finish
// (newest_writer) is inclusive. Iteration goes from old to new.
Writer* w = leader;
while (w != newest_writer) {
  w = w-&gt;link_newer;
.........................................
  w-&gt;write_group = write_group;
  size += batch_size;
  write_group-&gt;last_writer = w;
  write_group-&gt;size++;
}
..............................
}

这里注意到遍历是通过link_newer进行的，之所以这样做是相当于在写入WAL之前，对于当前leader的Write 做一次snapshot(通过CreateMissingNewerLinks函数).

void WriteThread::CreateMissingNewerLinks(Writer* head) {
while (true) {
  Writer* next = head-&gt;link_older;
  if (next == nullptr || next-&gt;link_newer != nullptr) {
    assert(next == nullptr || next-&gt;link_newer == head);
    break;
  }
  next-&gt;link_newer = head;
  head = next;
}
}

上述操作进行完毕之后，进入写WAL操作,最终会把这个write_group打包成一个writeBatch(通过MergeBatch函数)进行写入.

   if (w.ShouldWriteToWAL()) {
...............................
      w.status = WriteToWAL(wal_write_group, log_writer, log_used,
                            need_log_sync, need_log_dir_sync, current_sequence);
    }

当当前的leader将它自己与它的follow写入之后，此时它将需要写入memtable,那么此时之前还阻塞的Writer，分为两种情况第一种是已经被当前的leader打包写入到WAL，这些writer(包括leader自己)需要将他们链接到memtable writer list.还有一种情况，那就是还没有写入WAL的，此时这类writer则需要选择一个leader然后继续写入WAL.

void WriteThread::ExitAsBatchGroupLeader(WriteGroup&amp; write_group,
                                       Status status) {
Writer* leader = write_group.leader;
Writer* last_writer = write_group.last_writer;
assert(leader-&gt;link_older == nullptr);
.....................................

if (enable_pipelined_write_) {
  // Notify writers don't write to memtable to exit.
......................................
  // Link the ramaining of the group to memtable writer list.
  if (write_group.size &gt; 0) {
    if (LinkGroup(write_group, &amp;newest_memtable_writer_)) {
      // The leader can now be different from current writer.
      SetState(write_group.leader, STATE_MEMTABLE_WRITER_LEADER);
    }
  }
  // Reset newest_writer_ and wake up the next leader.
  Writer* newest_writer = last_writer;
  if (!newest_writer_.compare_exchange_strong(newest_writer, nullptr)) {
    Writer* next_leader = newest_writer;
    while (next_leader-&gt;link_older != last_writer) {
      next_leader = next_leader-&gt;link_older;
      assert(next_leader != nullptr);
    }
    next_leader-&gt;link_older = nullptr;
    SetState(next_leader, STATE_GROUP_LEADER);
  }
  AwaitState(leader, STATE_MEMTABLE_WRITER_LEADER |
                         STATE_PARALLEL_MEMTABLE_WRITER | STATE_COMPLETED,
             &amp;eabgl_ctx);
} else {
 .....................................
}
}

接下来我们来看写入memtable的操作，这里逻辑类似写入WAL,如果是leader的话，则依旧会创建一个group(WriteGroup),然后遍历需要写入memtable的writer,将他们都加入到group中(EnterAsMemTableWriter),然后则设置并发执行的大小，以及设置对应状态(LaunchParallelMemTableWriters).这里注意每次setstate就将会唤醒之前阻塞的Writer.

void WriteThread::LaunchParallelMemTableWriters(WriteGroup* write_group) {
assert(write_group != nullptr);
write_group-&gt;running.store(write_group-&gt;size);
for (auto w : *write_group) {
  SetState(w, STATE_PARALLEL_MEMTABLE_WRITER);
}
}

这里要注意，在构造memtable的group的时候，我们不需要创建link_newer，因为之前在写入WAL的时候，我们已经构造好link_newer,那么此时我们使用构造好的group也就是表示这个group中包含的都是已经写入到WAL的操作.

void WriteThread::EnterAsMemTableWriter(Writer* leader,
                                      WriteGroup* write_group) {
....................................

if (!allow_concurrent_memtable_write_ || !leader-&gt;batch-&gt;HasMerge()) {
 ....................................................
}

write_group-&gt;last_writer = last_writer;
write_group-&gt;last_sequence =
    last_writer-&gt;sequence + WriteBatchInternal::Count(last_writer-&gt;batch) - 1;
}

最后开始执行写入MemTable的操作，之前在写入WAL的时候被阻塞的所有Writer此时都会进入下面这个逻辑，此时也就意味着并发写入MemTable．

if (w.state == WriteThread::STATE_PARALLEL_MEMTABLE_WRITER) {
．．．．．．．．．．．．．．．．．．．．．．．．．
  w.status = WriteBatchInternal::InsertInto(
      &amp;w, w.sequence, &amp;column_family_memtables, &amp;flush_scheduler_,
      write_options.ignore_missing_column_families, 0 /*log_number*/, this,
      true /*concurrent_memtable_writes*/);
  if (write_thread_.CompleteParallelMemTableWriter(&amp;w)) {
    MemTableInsertStatusCheck(w.status);
    versions_-&gt;SetLastSequence(w.write_group-&gt;last_sequence);
    write_thread_.ExitAsMemTableWriter(&amp;w, *w.write_group);
  }
}

最后当当前group的所有Writer都写入MemTable之后，则将会调用ExitAsMemTableWriter来进行收尾工作.如果有新的memtable writer list需要处理，那么则唤醒对应的Writer,然后设置已经处理完毕的Writer的状态.

void WriteThread::ExitAsMemTableWriter(Writer* /*self*/,
                                     WriteGroup&amp; write_group) {
Writer* leader = write_group.leader;
Writer* last_writer = write_group.last_writer;

Writer* newest_writer = last_writer;
if (!newest_memtable_writer_.compare_exchange_strong(newest_writer,
                                                     nullptr)) {
  CreateMissingNewerLinks(newest_writer);
  Writer* next_leader = last_writer-&gt;link_newer;
  assert(next_leader != nullptr);
  next_leader-&gt;link_older = nullptr;
  SetState(next_leader, STATE_MEMTABLE_WRITER_LEADER);
}
Writer* w = leader;
while (true) {
  if (!write_group.status.ok()) {
    w-&gt;status = write_group.status;
  }
  Writer* next = w-&gt;link_newer;
  if (w != leader) {
    SetState(w, STATE_COMPLETED);
  }
  if (w == last_writer) {
    break;
  }
  w = next;
}
// Note that leader has to exit last, since it owns the write group.
SetState(leader, STATE_COMPLETED);
}

总结

我们可以看到在RocksDB中，WAL的写入始终是串行写入，而MemTable可以多线程并发写入，也就是说在系统压力到一定阶段的时候，写入WAL肯定会成为瓶颈.

文章来源：

Author：数据库内核月报
link：http://10.101.233.47:4000/monthly/2018/07/04/

下一篇：MySQL · 源码分析 · binlog crash recovery

上一篇：MongoDB · 引擎特性 · 事务实现解析

MySQL · RocksDB · 写入逻辑的实现

简介

实现

总结

添加我喜欢的博客

编辑我的关注

更多推荐博客

MySQL · RocksDB · 写入逻辑的实现

简介

实现

总结

添加我喜欢的博客

编辑 我的关注

更多 推荐博客

编辑我的关注

更多推荐博客