HBASE BufferedMutator 批量写入使用举例与源码解析

原创

大鹅

发布于 2021-06-04 15:55:24

2K0

发布于 2021-06-04 15:55:24

文章被收录于专栏：大鹅专栏：大数据到机器学习大鹅专栏：大数据到机器学习

toc

1. 基本介绍

BufferedMutator主要用来异步批量的将数据写入一个hbase表，就像Htable一样。通过Connection获取一个实例。

Map/reduce 任务是BufferedMutator的好的使用案例。Map/Reduce任务获益于batch操作，但是没有留出flush接口。BufferedMutator从Map/Reduce任务接受数据，会依据一些先验性的经验批量提交数据，比如puts堆积的数量，由于批量提交时异步的，所以M/R逻辑不会因为数据的batch提交而阻塞。Map/Reduce 批处理任务每个线程会有一个BufferedMutator。单个BufferedMutator也能够很高效用于大数据量的在线系统，来成批的写puts入hbase表。

2. BufferedMutator使用举例

这里分为以下两个批量写入场景

2.1 单次一张表批量写入

Configuration conf =  HBaseConfiguration.create();
 conf.set("hbase.zookeeper.quorum", "zookeeperHost");
final BufferedMutator.ExceptionListener listener = new BufferedMutator.ExceptionListener() {
@Override
public void onException(RetriesExhaustedWithDetailsException e, BufferedMutator mutator) {
 for (int i = 0; i < e.getNumExceptions(); i++) {
        LOG.info("Failed to sent put " + e.getRow(i) + ".");              }
        }
        };
 BufferedMutatorParams params = new BufferedMutatorParams(TABLE)
        .listener(listener);
 params.writeBufferSize(123123L);
        try {
        Connection conn = ConnectionFactory.createConnection(conf);
 BufferedMutator mutator = conn.getBufferedMutator(params);
 Put p = new Put(Bytes.toBytes("someRow"));
 p.addColumn(FAMILY, Bytes.toBytes("someQualifier"), Bytes.toBytes("some value"));
 mutator.mutate(p);
 mutator.close();
 conn.close();
 } catch (IOException e1) {
 // TODO Auto-generated catch block
 e1.printStackTrace();
 }

        }

多次多张表批量写入

可以使用一个Map保存多个Table的连接，这里使用的是线程安全的ConcurrentHashMap，如果是单线程的场景可以换成HashMap以提高效率。

private static Map<String, BufferedMutator> tableConnectionMgr = new ConcurrentHashMap<>();
private BufferedMutator getTableConnection(String tableName) throws IOException {
    if (tableConnectionMgr.get(tableName) != null) {
        return tableConnectionMgr.get(tableName);
    }
    Connection connection = ConnectionFactory.createConnection(config);
    BufferedMutator table = connection.getBufferedMutator(TableName.valueOf(tableName));
    tableConnectionMgr.put(tableName, table);
    log.info("hbase table: {} connect established!", tableName);
    return tableConnectionMgr.get(tableName);
}

3 源码介绍

3.1 主要类介绍

BufferedMutatorParams

实例化一个BufferedMutator所需要的参数。

主要参数TableName(表名)，writeBufferSize(写缓存大小)，maxKeyValueSize(最大key-value大小)，ExecutorService(执行线程池)，ExceptionListener(监听BufferedMutator的异常)。

BufferedMutatorImpl

用来和hbase表交互，类似于Htable，但是意味着批量，异步的puts。通过HConnectionImplementation获得实例，具体方法如下：

public BufferedMutator getBufferedMutator(BufferedMutatorParams params) {
 if (params.getTableName() == null) {
 throw new IllegalArgumentException("TableName cannot be null.");
 }
 if (params.getPool() == null) {
    params.pool(HTable.getDefaultExecutor(getConfiguration()));
 }
 if (params.getWriteBufferSize() == BufferedMutatorParams.UNSET) {
    params.writeBufferSize(connectionConfig.getWriteBufferSize());
 }
 if (params.getMaxKeyValueSize() == BufferedMutatorParams.UNSET) {
    params.maxKeyValueSize(connectionConfig.getMaxKeyValueSize());
 }
 return new BufferedMutatorImpl(this, rpcCallerFactory, rpcControllerFactory, params);
}

AsyncProcess

AsyncProcess内部维护的有一个线程池，我们的操作会被封装成runnable，然后扔到线程池里执行。这个过程是异步的，直到任务数达到最大值。

HConnectionImplementation

一个集群的链接。通过它可以找到master，定位到regions的分布，保持locations的缓存，并指导如何校准localtions信息。

3.2 源码过程

3.2.1 BufferedMutator构建的过程

首先是要构建一个HBaseConfiguration

Configuration conf =  HBaseConfiguration.create();  conf.set("hbase.zookeeper.quorum", "zookeeperHost");

接着是构建BufferedMutatorParams

```java

final BufferedMutator.ExceptionListener listener = new BufferedMutator.ExceptionListener() {

 @Override

 public void onException(RetriesExhaustedWithDetailsException e, BufferedMutator mutator) {

 for (int i = 0; i < e.getNumExceptions(); i++) {

 LOG.info("Failed to sent put " + e.getRow(i) + ".");

};

BufferedMutatorParams params = new BufferedMutatorParams(TABLE)

    .listener(listener);

params.writeBufferSize(123);

```

最后构建HConnection

Connection conn = ConnectionFactory.*createConnection*(getConf())

最后构建BufferMutator

BufferedMutator mutator = conn.getBufferedMutator(params)

3.2.2 数据发送的过程

构建put或者Listput
调用BufferedMutator.mutate方法
刷写到hbase

> 刷写到hbase三种方法：
>

> 一，显式调用BufferedMutator.flush
>

> 二，发送结束的时候调用BufferedMutator.close
>

> 三，它根据当前缓存大于了设置的写缓存大小

while (undealtMutationCount.get() != 0  && currentWriteBufferSize.get() > writeBufferSize) {   backgroundFlushCommits(false); }

最终都是调用的backgroundFlushCommits方法。

rpc的过程

入口是backgroundFlushCommits方法。Ap是AsyncProcess的实例。

ap.submit(tableName, taker, true, null, false);

首先是构建了一个HashMap，可以通过server找到该server上我们需要的region

```java

//可以根据我们的server找到要发送到该server的actions

Map<ServerName, MultiAction<Row>> actionsByServer =

 new HashMap<ServerName, MultiAction<Row>>();

```

获取所有的region信息，所有region的副本都被包括在内

```java

RegionLocations locs = connection.locateRegion(

    tableName, r.getRow(), true, true, RegionReplicaUtil.DEFAULT_REPLICA_ID);

```

获取默认的region信息此时一个region只会返回一个默认id指定的位置。

```java

loc = locs.getDefaultRegionLocation();

```

将row操作转变为action，并加入actionsByServer

```java

//可以操作将row操作变为Action

Action<Row> action = new Action<Row>(r, ++posInList);

setNonce(ng, r, action);

retainedActions.add(action);

// TODO: replica-get is not supported on this path

byte[] regionName = loc.getRegionInfo().getRegionName();

addAction(loc.getServerName(), regionName, action, actionsByServer, nonceGroup);

it.remove();

```

接着是

AsyncProcess.submitMultiActions

AsyncRequestFutureImpl<CResult>

.sendMultiAction(actionsByServer, 1, null, false);

内部主要是根据server，获取MultiAction，然后构建Runnable

```java

for (Map.Entry<ServerName, MultiAction<Row>> e : actionsByServer.entrySet()) {

  ServerName server = e.getKey();

 MultiAction<Row> multiAction = e.getValue();

 Collection<? extends Runnable> runnables = getNewMultiActionRunnable(server, multiAction,

 numAttempt);

 // make sure we correctly count the number of runnables before we try to reuse the send

  // thread, in case we had to split the request into different runnables because of backoff

 if (runnables.size() > actionsRemaining) {

    actionsRemaining = runnables.size();

```

然后，遍历执行Runnable

```java

for (Runnable runnable : runnables) {

 if ((--actionsRemaining == 0) && reuseThread

      && numAttempt % HConstants.DEFAULT_HBASE_CLIENT_RETRIES_NUMBER != 0) {

    runnable.run();

 } else {

 try {

 pool.submit(runnable);

```

Runnable的构建及Run方法

主要是进入getNewMultiActionRunnable

```java

List<Runnable> toReturn = new ArrayList<Runnable>(actions.size());

for (DelayingRunner runner : actions.values()) {

  incTaskCounters(runner.getActions().getRegions(), server);

 String traceText = "AsyncProcess.sendMultiAction";

 Runnable runnable = createSingleServerRequest(runner.getActions(), numAttempt, server, callsInProgress);

 // use a delay runner only if we need to sleep for some time

 if (runner.getSleepTime() > 0) {

    runner.setRunner(runnable);

 traceText = "AsyncProcess.clientBackoff.sendMultiAction";

 runnable = runner;

    if (connection.getConnectionMetrics() != null) {

 connection.getConnectionMetrics().incrDelayRunners();

 connection.getConnectionMetrics().updateDelayInterval(runner.getSleepTime());

  } else {

 if (connection.getConnectionMetrics() != null) {

 connection.getConnectionMetrics().incrNormalRunners();

  runnable = Trace.wrap(traceText, runnable);

 toReturn.add(runnable);

```

进入SingleServerRequestRunnable，分析其Run方法

```java

// setup the callable based on the actions, if we don't have one already from the request

if (callable == null) {

  callable = createCallable(server, tableName, multiAction);

RpcRetryingCaller<MultiResponse> caller = createCaller(callable, rpcTimeout);

try {

 if (callsInProgress != null) {

 callsInProgress.add(callable);

  res = caller.callWithoutRetries(callable, operationTimeout);

```

然后是RpcRetryingCaller中调用了MultiServerCallable的call方法，主要是构建请求，调用RPC。这就进入了服务端也即RSRpcServices的mutil方法。

```java

responseProto = getStub().multi(controller, requestProto);

```

3.2.3 HRegionserver端处理

RSRpcServices是服务端，本文对应的服务端实现是RSRpcServices.mutli。

if (request.hasCondition()) {
  Condition condition = request.getCondition();
  byte[] row = condition.getRow().toByteArray();
  byte[] family = condition.getFamily().toByteArray();
  byte[] qualifier = condition.getQualifier().toByteArray();
 CompareOp compareOp = CompareOp.valueOf(condition.getCompareType().name());
 ByteArrayComparable comparator =
      ProtobufUtil.toComparator(condition.getComparator());
 processed = checkAndRowMutate(region, regionAction.getActionList(),
 cellScanner, row, family, qualifier, compareOp,
 comparator, regionActionResultBuilder);
} else {
  mutateRows(region, regionAction.getActionList(), cellScanner,
 regionActionResultBuilder);
 processed = Boolean.TRUE;
}

根据条件进入checkAndRowMutate或者mutateRows。

根据类型做不同的操作，然后正式进入执行操作

MutationType type = action.getMutation().getMutateType();
  if (rm == null) {
    rm = new RowMutations(action.getMutation().getRow().toByteArray());
 }
 switch (type) {
 case PUT:
      rm.add(ProtobufUtil.toPut(action.getMutation(), cellScanner));
      break;
    case DELETE:
      rm.add(ProtobufUtil.toDelete(action.getMutation(), cellScanner));
      break;
    default:
 throw new DoNotRetryIOException("Atomic put and/or delete only, not " + type.name());
 }
 // To unify the response format with doNonAtomicRegionMutation and read through client's
  // AsyncProcess we have to add an empty result instance per operation
 resultOrExceptionOrBuilder.clear();
 resultOrExceptionOrBuilder.setIndex(i++);
 builder.addResultOrException(
      resultOrExceptionOrBuilder.build());
}
region.mutateRow(rm);

HRegion.mutateRow方法

HRegion.mutateRowsWithLocks

public void mutateRowsWithLocks(Collection<Mutation> mutations,
 Collection<byte[]> rowsToLock) throws IOException {
  mutateRowsWithLocks(mutations, rowsToLock, HConstants.NO_NONCE, HConstants.NO_NONCE);
}
public void mutateRowsWithLocks(Collection<Mutation> mutations,
 Collection<byte[]> rowsToLock, long nonceGroup, long nonce) throws IOException {
  MultiRowMutationProcessor proc = new MultiRowMutationProcessor(mutations, rowsToLock);
 processRowsWithLocks(proc, -1, nonceGroup, nonce);
}

具体处理的过程，可以自行去看了，源码注释条例很清晰。

4. 总结

Hbase的JAVA API客户端，写操作有三种实现：

HTablePool

源码请看hbase权威指南。

![](https://img-blog.csdnimg.cn/20190914124230128.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FzZDEzNjkxMg==,size_16,color_FFFFFF,t_70)

HConnection

这种方式要自己实现一个线程池。

```java

Connection conn = ConnectionFactory.createConnection(conf);

TableName tabName=  TableName.valueOf("tableName");

Table table=conn.getTable(tabName);

```

BufferedMutator

建议put操作采用这种方式。

批量，异步puts操作。

5. Ref

https://cloud.tencent.com/developer/article/1032502
hbase权威指南

原创声明：本文系作者授权腾讯云开发者社区发表，未经许可，不得转载。

如有侵权，请联系 cloudcommunity@tencent.com 删除。

hbase

云服务器

弹性 MapReduce

大数据处理套件 TBDS