• 设为首页
  • 点击收藏
  • 手机版
    手机扫一扫访问
    迪恩网络手机版
  • 关注官方公众号
    微信扫一扫关注
    公众号
登陆 注册

Java SplitLogTask类代码示例

原作者: [db:作者] 来自: [db:来源] 收藏 邀请

本文整理汇总了Java中org.apache.hadoop.hbase.SplitLogTask类的典型用法代码示例。如果您正苦于以下问题：Java SplitLogTask类的具体用法？Java SplitLogTask怎么用？Java SplitLogTask使用的例子？那么恭喜您, 这里精选的类代码示例或许可以为您提供帮助。



SplitLogTask类属于org.apache.hadoop.hbase包，在下文中一共展示了SplitLogTask类的20个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于我们的系统推荐出更棒的Java代码示例。

示例1: rescan

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
/**
 * signal the workers that a task was resubmitted by creating the RESCAN node.
 */
private void rescan(long retries) {
  // The RESCAN node will be deleted almost immediately by the
  // SplitLogManager as soon as it is created because it is being
  // created in the DONE state. This behavior prevents a buildup
  // of RESCAN nodes. But there is also a chance that a SplitLogWorker
  // might miss the watch-trigger that creation of RESCAN node provides.
  // Since the TimeoutMonitor will keep resubmitting UNASSIGNED tasks
  // therefore this behavior is safe.
  SplitLogTask slt = new SplitLogTask.Done(this.details.getServerName(), getRecoveryMode());
  this.watcher
      .getRecoverableZooKeeper()
      .getZooKeeper()
      .create(ZKSplitLog.getRescanNode(watcher), slt.toByteArray(), Ids.OPEN_ACL_UNSAFE,
        CreateMode.EPHEMERAL_SEQUENTIAL, new CreateRescanAsyncCallback(), Long.valueOf(retries));
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:19，代码来源:ZKSplitLogManagerCoordination.java


示例2: testOrphanTaskAcquisition

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testOrphanTaskAcquisition() throws Exception {
  LOG.info("TestOrphanTaskAcquisition");

  String tasknode = ZKSplitLog.getEncodedNodeName(zkw, "orphan/test/slash");
  SplitLogTask slt = new SplitLogTask.Owned(DUMMY_MASTER, this.mode);
  zkw.getRecoverableZooKeeper().create(tasknode, slt.toByteArray(), Ids.OPEN_ACL_UNSAFE,
      CreateMode.PERSISTENT);

  slm = new SplitLogManager(ds, conf, stopper, master, DUMMY_MASTER);
  waitForCounter(tot_mgr_orphan_task_acquired, 0, 1, to/2);
  Task task = slm.findOrCreateOrphanTask(tasknode);
  assertTrue(task.isOrphan());
  waitForCounter(tot_mgr_heartbeat, 0, 1, to/2);
  assertFalse(task.isUnassigned());
  long curt = System.currentTimeMillis();
  assertTrue((task.last_update <= curt) &&
      (task.last_update > (curt - 1000)));
  LOG.info("waiting for manager to resubmit the orphan task");
  waitForCounter(tot_mgr_resubmit, 0, 1, to + to/2);
  assertTrue(task.isUnassigned());
  waitForCounter(tot_mgr_rescan, 0, 1, to + to/2);
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:24，代码来源:TestSplitLogManager.java


示例3: testTaskDone

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testTaskDone() throws Exception {
  LOG.info("TestTaskDone - cleanup task node once in DONE state");

  slm = new SplitLogManager(ds, conf, stopper, master, DUMMY_MASTER);
  TaskBatch batch = new TaskBatch();
  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  SplitLogTask slt = new SplitLogTask.Done(worker1, this.mode);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  synchronized (batch) {
    while (batch.installed != batch.done) {
      batch.wait();
    }
  }
  waitForCounter(tot_mgr_task_deleted, 0, 1, to/2);
  assertTrue(ZKUtil.checkExists(zkw, tasknode) == -1);
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:19，代码来源:TestSplitLogManager.java


示例4: testTaskErr

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testTaskErr() throws Exception {
  LOG.info("TestTaskErr - cleanup task node once in ERR state");

  conf.setInt("hbase.splitlog.max.resubmit", 0);
  slm = new SplitLogManager(ds, conf, stopper, master, DUMMY_MASTER);
  TaskBatch batch = new TaskBatch();

  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  SplitLogTask slt = new SplitLogTask.Err(worker1, this.mode);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());

  synchronized (batch) {
    while (batch.installed != batch.error) {
      batch.wait();
    }
  }
  waitForCounter(tot_mgr_task_deleted, 0, 1, to/2);
  assertTrue(ZKUtil.checkExists(zkw, tasknode) == -1);
  conf.setInt("hbase.splitlog.max.resubmit", ZKSplitLogManagerCoordination.DEFAULT_MAX_RESUBMIT);
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:23，代码来源:TestSplitLogManager.java


示例5: testTaskResigned

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testTaskResigned() throws Exception {
  LOG.info("TestTaskResigned - resubmit task node once in RESIGNED state");
  assertEquals(tot_mgr_resubmit.get(), 0);
  slm = new SplitLogManager(ds, conf, stopper, master, DUMMY_MASTER);
  assertEquals(tot_mgr_resubmit.get(), 0);
  TaskBatch batch = new TaskBatch();
  String tasknode = submitTaskAndWait(batch, "foo/1");
  assertEquals(tot_mgr_resubmit.get(), 0);
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  assertEquals(tot_mgr_resubmit.get(), 0);
  SplitLogTask slt = new SplitLogTask.Resigned(worker1, this.mode);
  assertEquals(tot_mgr_resubmit.get(), 0);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  int version = ZKUtil.checkExists(zkw, tasknode);
  // Could be small race here.
  if (tot_mgr_resubmit.get() == 0) {
    waitForCounter(tot_mgr_resubmit, 0, 1, to/2);
  }
  assertEquals(tot_mgr_resubmit.get(), 1);

  byte[] taskstate = ZKUtil.getData(zkw, tasknode);
  slt = SplitLogTask.parseFrom(taskstate);
  assertTrue(slt.isUnassigned(DUMMY_MASTER));
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:26，代码来源:TestSplitLogManager.java


示例6: testWorkerCrash

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testWorkerCrash() throws Exception {
  slm = new SplitLogManager(ds, conf, stopper, master, DUMMY_MASTER);
  TaskBatch batch = new TaskBatch();

  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");

  SplitLogTask slt = new SplitLogTask.Owned(worker1, this.mode);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  if (tot_mgr_heartbeat.get() == 0) waitForCounter(tot_mgr_heartbeat, 0, 1, to/2);

  // Not yet resubmitted.
  Assert.assertEquals(0, tot_mgr_resubmit.get());

  // This server becomes dead
  Mockito.when(sm.isServerOnline(worker1)).thenReturn(false);

  Thread.sleep(1300); // The timeout checker is done every 1000 ms (hardcoded).

  // It has been resubmitted
  Assert.assertEquals(1, tot_mgr_resubmit.get());
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:24，代码来源:TestSplitLogManager.java


示例7: testGetPreviousRecoveryMode

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Ignore("DLR is broken by HBASE-12751") @Test(timeout=60000)
public void testGetPreviousRecoveryMode() throws Exception {
  LOG.info("testGetPreviousRecoveryMode");
  SplitLogCounters.resetCounters();
  // Not actually enabling DLR for the cluster, just for the ZkCoordinatedStateManager to use.
  // The test is just manipulating ZK manually anyways.
  conf.setBoolean(HConstants.DISTRIBUTED_LOG_REPLAY_KEY, true);

  zkw.getRecoverableZooKeeper().create(ZKSplitLog.getEncodedNodeName(zkw, "testRecovery"),
    new SplitLogTask.Unassigned(
      ServerName.valueOf("mgr,1,1"), RecoveryMode.LOG_SPLITTING).toByteArray(),
      Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);

  slm = new SplitLogManager(ds, conf, stopper, master, DUMMY_MASTER);
  LOG.info("Mode1=" + slm.getRecoveryMode());
  assertTrue(slm.isLogSplitting());
  zkw.getRecoverableZooKeeper().delete(ZKSplitLog.getEncodedNodeName(zkw, "testRecovery"), -1);
  LOG.info("Mode2=" + slm.getRecoveryMode());
  slm.setRecoveryMode(false);
  LOG.info("Mode3=" + slm.getRecoveryMode());
  assertTrue("Mode4=" + slm.getRecoveryMode(), slm.isLogReplaying());
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:23，代码来源:TestSplitLogManager.java


示例8: testAcquireTaskAtStartup

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test(timeout=60000)
public void testAcquireTaskAtStartup() throws Exception {
  LOG.info("testAcquireTaskAtStartup");
  SplitLogCounters.resetCounters();
  final String TATAS = "tatas";
  final ServerName RS = ServerName.valueOf("rs,1,1");
  RegionServerServices mockedRS = getRegionServer(RS);
  zkw.getRecoverableZooKeeper().create(ZKSplitLog.getEncodedNodeName(zkw, TATAS),
    new SplitLogTask.Unassigned(ServerName.valueOf("mgr,1,1"), this.mode).toByteArray(),
      Ids.OPEN_ACL_UNSAFE,
      CreateMode.PERSISTENT);

  SplitLogWorker slw =
      new SplitLogWorker(ds, TEST_UTIL.getConfiguration(), mockedRS, neverEndingTask);
  slw.start();
  try {
    waitForCounter(SplitLogCounters.tot_wkr_task_acquired, 0, 1, WAIT_TIME);
    byte [] bytes = ZKUtil.getData(zkw, ZKSplitLog.getEncodedNodeName(zkw, TATAS));
    SplitLogTask slt = SplitLogTask.parseFrom(bytes);
    assertTrue(slt.isOwned(RS));
  } finally {
   stopSplitLogWorker(slw);
  }
}
开发者ID:fengchen8086，项目名称:ditb，代码行数:25，代码来源:TestSplitLogWorker.java


示例9: testGetPreviousRecoveryMode

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test(timeout=60000)
public void testGetPreviousRecoveryMode() throws Exception {
  LOG.info("testGetPreviousRecoveryMode");
  SplitLogCounters.resetCounters();
  Configuration testConf = HBaseConfiguration.create(TEST_UTIL.getConfiguration());
  testConf.setBoolean(HConstants.DISTRIBUTED_LOG_REPLAY_KEY, true);
  ds = new DummyServer(zkw, testConf);

  zkw.getRecoverableZooKeeper().create(ZKSplitLog.getEncodedNodeName(zkw, "testRecovery"),
    new SplitLogTask.Unassigned(
      ServerName.valueOf("mgr,1,1"), RecoveryMode.LOG_SPLITTING).toByteArray(),
      Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);

  slm = new SplitLogManager(ds, testConf, stopper, master, DUMMY_MASTER);
  LOG.info("Mode1=" + slm.getRecoveryMode());
  assertTrue(slm.isLogSplitting());
  zkw.getRecoverableZooKeeper().delete(ZKSplitLog.getEncodedNodeName(zkw, "testRecovery"), -1);
  LOG.info("Mode2=" + slm.getRecoveryMode());
  slm.setRecoveryMode(false);
  LOG.info("Mode3=" + slm.getRecoveryMode());
  assertTrue("Mode4=" + slm.getRecoveryMode(), slm.isLogReplaying());
}
开发者ID:grokcoder，项目名称:pbase，代码行数:23，代码来源:TestSplitLogManager.java


示例10: createRescanNode

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
/**
 * signal the workers that a task was resubmitted by creating the
 * RESCAN node.
 * @throws KeeperException
 */
private void createRescanNode(long retries) {
  // The RESCAN node will be deleted almost immediately by the
  // SplitLogManager as soon as it is created because it is being
  // created in the DONE state. This behavior prevents a buildup
  // of RESCAN nodes. But there is also a chance that a SplitLogWorker
  // might miss the watch-trigger that creation of RESCAN node provides.
  // Since the TimeoutMonitor will keep resubmitting UNASSIGNED tasks
  // therefore this behavior is safe.
  lastTaskCreateTime = EnvironmentEdgeManager.currentTimeMillis();
  SplitLogTask slt = new SplitLogTask.Done(this.serverName);
  this.watcher.getRecoverableZooKeeper().getZooKeeper().
    create(ZKSplitLog.getRescanNode(watcher), slt.toByteArray(),
      Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL_SEQUENTIAL,
      new CreateRescanAsyncCallback(), Long.valueOf(retries));
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:21，代码来源:SplitLogManager.java


示例11: testOrphanTaskAcquisition

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test
public void testOrphanTaskAcquisition() throws Exception {
  LOG.info("TestOrphanTaskAcquisition");

  String tasknode = ZKSplitLog.getEncodedNodeName(zkw, "orphan/test/slash");
  SplitLogTask slt = new SplitLogTask.Owned(DUMMY_MASTER);
  zkw.getRecoverableZooKeeper().create(tasknode, slt.toByteArray(), Ids.OPEN_ACL_UNSAFE,
      CreateMode.PERSISTENT);

  slm = new SplitLogManager(zkw, conf, stopper, master, DUMMY_MASTER);
  waitForCounter(tot_mgr_orphan_task_acquired, 0, 1, to/2);
  Task task = slm.findOrCreateOrphanTask(tasknode);
  assertTrue(task.isOrphan());
  waitForCounter(tot_mgr_heartbeat, 0, 1, to/2);
  assertFalse(task.isUnassigned());
  long curt = System.currentTimeMillis();
  assertTrue((task.last_update <= curt) &&
      (task.last_update > (curt - 1000)));
  LOG.info("waiting for manager to resubmit the orphan task");
  waitForCounter(tot_mgr_resubmit, 0, 1, to + to/2);
  assertTrue(task.isUnassigned());
  waitForCounter(tot_mgr_rescan, 0, 1, to + to/2);
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:24，代码来源:TestSplitLogManager.java


示例12: testTaskDone

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test
public void testTaskDone() throws Exception {
  LOG.info("TestTaskDone - cleanup task node once in DONE state");

  slm = new SplitLogManager(zkw, conf, stopper, master, DUMMY_MASTER);
  TaskBatch batch = new TaskBatch();
  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  SplitLogTask slt = new SplitLogTask.Done(worker1);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  synchronized (batch) {
    while (batch.installed != batch.done) {
      batch.wait();
    }
  }
  waitForCounter(tot_mgr_task_deleted, 0, 1, to/2);
  assertTrue(ZKUtil.checkExists(zkw, tasknode) == -1);
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:19，代码来源:TestSplitLogManager.java


示例13: testTaskErr

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test
public void testTaskErr() throws Exception {
  LOG.info("TestTaskErr - cleanup task node once in ERR state");

  conf.setInt("hbase.splitlog.max.resubmit", 0);
  slm = new SplitLogManager(zkw, conf, stopper, master, DUMMY_MASTER);
  TaskBatch batch = new TaskBatch();

  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  SplitLogTask slt = new SplitLogTask.Err(worker1);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());

  synchronized (batch) {
    while (batch.installed != batch.error) {
      batch.wait();
    }
  }
  waitForCounter(tot_mgr_task_deleted, 0, 1, to/2);
  assertTrue(ZKUtil.checkExists(zkw, tasknode) == -1);
  conf.setInt("hbase.splitlog.max.resubmit", SplitLogManager.DEFAULT_MAX_RESUBMIT);
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:23，代码来源:TestSplitLogManager.java


示例14: testTaskResigned

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test
public void testTaskResigned() throws Exception {
  LOG.info("TestTaskResigned - resubmit task node once in RESIGNED state");
  assertEquals(tot_mgr_resubmit.get(), 0);
  slm = new SplitLogManager(zkw, conf, stopper, master, DUMMY_MASTER);
  assertEquals(tot_mgr_resubmit.get(), 0);
  TaskBatch batch = new TaskBatch();
  String tasknode = submitTaskAndWait(batch, "foo/1");
  assertEquals(tot_mgr_resubmit.get(), 0);
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  assertEquals(tot_mgr_resubmit.get(), 0);
  SplitLogTask slt = new SplitLogTask.Resigned(worker1);
  assertEquals(tot_mgr_resubmit.get(), 0);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  int version = ZKUtil.checkExists(zkw, tasknode);
  // Could be small race here.
  if (tot_mgr_resubmit.get() == 0) {
    waitForCounter(tot_mgr_resubmit, 0, 1, to/2);
  }
  assertEquals(tot_mgr_resubmit.get(), 1);

  byte[] taskstate = ZKUtil.getData(zkw, tasknode);
  slt = SplitLogTask.parseFrom(taskstate);
  assertTrue(slt.isUnassigned(DUMMY_MASTER));
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:26，代码来源:TestSplitLogManager.java


示例15: testWorkerCrash

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test
public void testWorkerCrash() throws Exception {
  slm = new SplitLogManager(zkw, conf, stopper, master, DUMMY_MASTER);
  TaskBatch batch = new TaskBatch();

  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");

  SplitLogTask slt = new SplitLogTask.Owned(worker1);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  if (tot_mgr_heartbeat.get() == 0) waitForCounter(tot_mgr_heartbeat, 0, 1, to/2);

  // Not yet resubmitted.
  Assert.assertEquals(0, tot_mgr_resubmit.get());

  // This server becomes dead
  Mockito.when(sm.isServerOnline(worker1)).thenReturn(false);

  Thread.sleep(1300); // The timeout checker is done every 1000 ms (hardcoded).

  // It has been resubmitted
  Assert.assertEquals(1, tot_mgr_resubmit.get());
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:24，代码来源:TestSplitLogManager.java


示例16: testAcquireTaskAtStartup

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test(timeout=60000)
public void testAcquireTaskAtStartup() throws Exception {
  LOG.info("testAcquireTaskAtStartup");
  SplitLogCounters.resetCounters();
  final String TATAS = "tatas";
  final ServerName RS = ServerName.valueOf("rs,1,1");
  RegionServerServices mockedRS = getRegionServer(RS);
  zkw.getRecoverableZooKeeper().create(ZKSplitLog.getEncodedNodeName(zkw, TATAS),
    new SplitLogTask.Unassigned(ServerName.valueOf("mgr,1,1")).toByteArray(), Ids.OPEN_ACL_UNSAFE,
      CreateMode.PERSISTENT);

  SplitLogWorker slw =
      new SplitLogWorker(zkw, TEST_UTIL.getConfiguration(), mockedRS, neverEndingTask);
  slw.start();
  try {
    waitForCounter(SplitLogCounters.tot_wkr_task_acquired, 0, 1, WAIT_TIME);
    byte [] bytes = ZKUtil.getData(zkw, ZKSplitLog.getEncodedNodeName(zkw, TATAS));
    SplitLogTask slt = SplitLogTask.parseFrom(bytes);
    assertTrue(slt.isOwned(RS));
  } finally {
    stopSplitLogWorker(slw);
  }
}
开发者ID:tenggyut，项目名称:HIndex，代码行数:24，代码来源:TestSplitLogWorker.java


示例17: rescan

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
/**
 * signal the workers that a task was resubmitted by creating the RESCAN node.
 */
private void rescan(long retries) {
  // The RESCAN node will be deleted almost immediately by the
  // SplitLogManager as soon as it is created because it is being
  // created in the DONE state. This behavior prevents a buildup
  // of RESCAN nodes. But there is also a chance that a SplitLogWorker
  // might miss the watch-trigger that creation of RESCAN node provides.
  // Since the TimeoutMonitor will keep resubmitting UNASSIGNED tasks
  // therefore this behavior is safe.
  SplitLogTask slt = new SplitLogTask.Done(this.details.getServerName());
  this.watcher
      .getRecoverableZooKeeper()
      .getZooKeeper()
      .create(ZKSplitLog.getRescanNode(watcher), slt.toByteArray(), Ids.OPEN_ACL_UNSAFE,
        CreateMode.EPHEMERAL_SEQUENTIAL, new CreateRescanAsyncCallback(), Long.valueOf(retries));
}
开发者ID:apache，项目名称:hbase，代码行数:19，代码来源:ZKSplitLogManagerCoordination.java


示例18: testOrphanTaskAcquisition

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testOrphanTaskAcquisition() throws Exception {
  LOG.info("TestOrphanTaskAcquisition");

  String tasknode = ZKSplitLog.getEncodedNodeName(zkw, "orphan/test/slash");
  SplitLogTask slt = new SplitLogTask.Owned(master.getServerName());
  zkw.getRecoverableZooKeeper().create(tasknode, slt.toByteArray(), Ids.OPEN_ACL_UNSAFE,
      CreateMode.PERSISTENT);

  slm = new SplitLogManager(master, conf);
  waitForCounter(tot_mgr_orphan_task_acquired, 0, 1, to/2);
  Task task = findOrCreateOrphanTask(tasknode);
  assertTrue(task.isOrphan());
  waitForCounter(tot_mgr_heartbeat, 0, 1, to/2);
  assertFalse(task.isUnassigned());
  long curt = System.currentTimeMillis();
  assertTrue((task.last_update <= curt) &&
      (task.last_update > (curt - 1000)));
  LOG.info("waiting for manager to resubmit the orphan task");
  waitForCounter(tot_mgr_resubmit, 0, 1, to + to/2);
  assertTrue(task.isUnassigned());
  waitForCounter(tot_mgr_rescan, 0, 1, to + to/2);
}
开发者ID:apache，项目名称:hbase，代码行数:24，代码来源:TestSplitLogManager.java


示例19: testTaskDone

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testTaskDone() throws Exception {
  LOG.info("TestTaskDone - cleanup task node once in DONE state");

  slm = new SplitLogManager(master, conf);
  TaskBatch batch = new TaskBatch();
  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  SplitLogTask slt = new SplitLogTask.Done(worker1);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());
  synchronized (batch) {
    while (batch.installed != batch.done) {
      batch.wait();
    }
  }
  waitForCounter(tot_mgr_task_deleted, 0, 1, to/2);
  assertTrue(ZKUtil.checkExists(zkw, tasknode) == -1);
}
开发者ID:apache，项目名称:hbase，代码行数:19，代码来源:TestSplitLogManager.java


示例20: testTaskErr

import org.apache.hadoop.hbase.SplitLogTask; //导入依赖的package包/类
@Test (timeout=180000)
public void testTaskErr() throws Exception {
  LOG.info("TestTaskErr - cleanup task node once in ERR state");

  conf.setInt("hbase.splitlog.max.resubmit", 0);
  slm = new SplitLogManager(master, conf);
  TaskBatch batch = new TaskBatch();

  String tasknode = submitTaskAndWait(batch, "foo/1");
  final ServerName worker1 = ServerName.valueOf("worker1,1,1");
  SplitLogTask slt = new SplitLogTask.Err(worker1);
  ZKUtil.setData(zkw, tasknode, slt.toByteArray());

  synchronized (batch) {
    while (batch.installed != batch.error) {
      batch.wait();
    }
  }
  waitForCounter(tot_mgr_task_deleted, 0, 1, to/2);
  assertTrue(ZKUtil.checkExists(zkw, tasknode) == -1);
  conf.setInt("hbase.splitlog.max.resubmit", ZKSplitLogManagerCoordination.DEFAULT_MAX_RESUBMIT);
}
开发者ID:apache，项目名称:hbase，代码行数:23，代码来源:TestSplitLogManager.java



注：本文中的org.apache.hadoop.hbase.SplitLogTask类示例整理自Github/MSDocs等源码及文档管理平台，相关代码片段筛选自各路编程大神贡献的开源项目，源码版权归原作者所有，传播和使用请参考对应项目的License；未经允许，请勿转载。

鲜花
握手
雷人
路过
鸡蛋
该文章已有0人参与评论

请发表评论

全部评论

专题导读
More+
上一篇：
Java Link类代码示例发布时间：2022-05-22
下一篇：
Java MethodInvocationReport类代码示例发布时间：2022-05-22
热门推荐
More+
热门话题
More+
阅读排行榜

扫描微信二维码

查看手机版网站

随时了解更新最新资讯

139-2527-9053

在线客服（服务时间 9:00～18:00）

在线QQ客服
地址：深圳市南山区西丽大学城创智工业园
电邮：jeky_zhao#qq.com
移动电话：139-2527-9053

Powered by 互联科技 X3.4© 2001-2213 极客世界.|Sitemap