HBase数据达到Memstore的过程

往HBase表中写入数据的主要流程。

1 RegionServer启动的初始化工作

启动RegionServer进程的入口类是HRegionServer，通过执行该类的main方法开始。

其简单流程如下：

//通过main方法的执行流程
//main()
new HRegionServerCommandLine(regionServerClass).doMain(args);

//doMain
int ret = ToolRunner.run(HBaseConfiguration.create(), this, args);

//ToolRunner.run, 启动的tool就是HRegionServerCommandLine对象
run(tool.getConf);

//HRegionServerCommandLine.run  HRegionServerCommandLine.start
//构造一个HRegionServer线程对象，并启动
HRegionServer hrs = HRegionServer.constructRegionServer(regionServerClass, conf);
hrs.start();
handleReportForDutyResponse(w);
startServices();
initializeMemStoreChunkCreator();

//这里定义chunksize的默认大小为2M
int chunkSize = conf.getInt(MemStoreLAB.CHUNK_SIZE_KEY, MemStoreLAB.CHUNK_SIZE_DEFAULT);
// init the chunkCreator
ChunkCreator.initialize(chunkSize, offheap, globalMemStoreSize, poolSizePercentage,
                        initialCountPercentage, this.hMemManager);

//initialize
instance = new ChunkCreator(chunkSize, offheap, globalMemStoreSize, poolSizePercentage,
            initialCountPercentage, heapMemoryManager,
            MemStoreLABImpl.INDEX_CHUNK_PERCENTAGE_DEFAULT);

//ChunkCreator构造方法
this.chunkSize = chunkSize; // in case pools are not allocated
initializePools(chunkSize, globalMemStoreSize, poolSizePercentage, indexChunkSizePercentage,
                initialCountPercentage, heapMemoryManager);

//initializePools
//在这里定义了dataChunk、indexChunk的size大小，默认为2M
this.dataChunksPool = initializePool("data", globalMemStoreSize,
            (1 - indexChunkSizePercentage) * poolSizePercentage,
            initialCountPercentage, chunkSize, heapMemoryManager);
// The index chunks pool is needed only when the index type is CCM.
// Since the pools are not created at all when the index type isn't CCM,
// we don't need to check it here.
this.indexChunksPool = initializePool("index", globalMemStoreSize,
                                      indexChunkSizePercentage * poolSizePercentage,
                                      initialCountPercentage, (int) (indexChunkSizePercentage * chunkSize),
                                      heapMemoryManager);

2 put操作的流程

当往HBase表中put数据时，最终会调用HRgion中的doMiniBatchMutate方法，接下来的调用流程如下所示。

//doMiniBatchMutate
//在此方法内，会先将修改写入到WAl中
// STEP 3. Build WAL edit
List<Pair<NonceKey, WALEdit>> walEdits = batchOp.buildWALEdits(miniBatchOp);

// STEP 4. Append the WALEdits to WAL and sync.
for(Iterator<Pair<NonceKey, WALEdit>> it = walEdits.iterator(); it.hasNext();) {
   
    Pair<NonceKey, WALEdit> nonceKeyWALEditPair = it.next();
    walEdit = nonceKeyWALEditPair.getSecond();
    NonceKey nonceKey = nonceKeyWALEditPair.getFirst();

    if (walEdit != null && !walEdit.isEmpty()) {
   
        writeEntry = doWALAppend(walEdit, batchOp.durability, batchOp.getClusterIds<