com.alibaba.wasp.zookeeper.ZKAssign类的使用及代码示例

x33g5p2x  于2022-02-05 转载在 其他  
字(12.8k)|赞(0)|评价(0)|浏览(118)

本文整理了Java中com.alibaba.wasp.zookeeper.ZKAssign类的一些代码示例，展示了ZKAssign类的具体用法。这些代码示例主要来源于Github/Stackoverflow/Maven等平台，是从一些精选项目中提取出来的代码，具有较强的参考意义，能在一定程度帮忙到你。ZKAssign类的具体详情如下：
包路径：com.alibaba.wasp.zookeeper.ZKAssign
类名称：ZKAssign

ZKAssign介绍

[英]Utility class for doing entityGroup assignment in ZooKeeper. This class extends stuff done in ZKUtil to cover specific assignment operations.

Contains only static methods and constants.

Used by both the FMaster and FServer.

All valid transitions outlined below:

FMASTER

  1. FMaster creates an unassigned node as OFFLINE. - Cluster startup and table enabling.
  2. FMaster forces an existing unassigned node to OFFLINE. - FServer failure. - Allows transitions from all states to OFFLINE.
  3. FMaster deletes an unassigned node that was in a OPENED state. - Normal entityGroup transitions. Besides cluster startup, no other deletions of unassigned nodes is allowed.
  4. FMaster deletes all unassigned nodes regardless of state. - Cluster startup before any assignment happens.

FSERVER

  1. FServer creates an unassigned node as CLOSING. - All entityGroups closes will do this in response to a CLOSE RPC from FMaster. - A node can never be transitioned to CLOSING, only created.
  2. FServer transitions an unassigned node from CLOSING to CLOSED. - Normal entityGroup closes. CAS operation.
  3. FServer transitions an unassigned node from OFFLINE to OPENING. - All entityGroup opens will do this in response to an OPEN RPC from the FMaster. - Normal entityGroup opens. CAS operation.
  4. FServer transitions an unassigned node from OPENING to OPENED. - Normal entityGroup opens. CAS operation.
    [中]用于在ZooKeeper中执行entityGroup分配的实用程序类。这个类扩展了在ZKUtil中完成的工作，以涵盖特定的赋值操作。
    只包含静态方法和常量。
    由FMaster和FServer使用。
    以下列出了所有有效的转换：
    大师
    1.FMaster将未分配的节点创建为脱机。-集群启动和表启用。
    1.FMaster强制现有未分配节点脱机。-F服务器故障。-允许从所有状态转换为脱机状态。
    1.FMaster删除处于打开状态的未分配节点。-正常的实体组转换。除了集群启动，不允许删除其他未分配的节点。
    1.无论状态如何，FMaster都会删除所有未分配的节点。-在任何分配发生之前启动群集。
    服务器
    1.FServer创建一个未分配的节点作为关闭。-所有entityGroups closes都将这样做，以响应FMaster发出的关闭RPC。-节点永远不能转换为关闭，只能创建。
    1.FServer将未分配的节点从关闭转换为关闭。-正常实体组关闭。CAS操作。
    1.FServer将未分配的节点从脱机转换为打开。-所有entityGroup打开时都会响应FMaster打开的RPC来执行此操作。-正常实体组打开。CAS操作。
    1.FServer将未分配的节点从打开状态转换为打开状态。-正常实体组打开。CAS操作。

代码示例

代码示例来源：origin: alibaba/wasp

/**
 * Creates a new unassigned node in the OFFLINE state for the specified
 * entityGroup.
 *
 * <p>
 * Does not transition nodes from other states. If a node already exists for
 * this entityGroup, a {@link org.apache.zookeeper.KeeperException.NodeExistsException} will be thrown.
 *
 * <p>
 * Sets a watcher on the unassigned entityGroup node if the method is
 * successful.
 *
 * <p>
 * This method should only be used during cluster startup and the enabling of
 * a table.
 *
 * @param zkw zk reference
 * @param entityGroup entityGroup to be created as offline
 * @param serverName server transition will happen on
 * @throws org.apache.zookeeper.KeeperException if unexpected zookeeper exception
 * @throws org.apache.zookeeper.KeeperException.NodeExistsException if node already exists
 */
public static void createNodeOffline(ZooKeeperWatcher zkw,
  EntityGroupInfo entityGroup, ServerName serverName) throws KeeperException,
  KeeperException.NodeExistsException {
 createNodeOffline(zkw, entityGroup, serverName,
   EventType.M_ZK_ENTITYGROUP_OFFLINE);
}

代码示例来源：origin: alibaba/wasp

/**
 * Deletes an existing unassigned node that is in the OFFLINE state for the
 * specified entityGroup.
 *
 * <p>
 * If a node does not already exist for this entityGroup, a
 * {@link org.apache.zookeeper.KeeperException.NoNodeException} will be thrown.
 *
 * <p>
 * No watcher is set whether this succeeds or not.
 *
 * <p>
 * Returns false if the node was not in the proper state but did exist.
 *
 * <p>
 * This method is used during master failover when the entityGroups on an RS
 * that has died are all set to OFFLINE before being processed.
 *
 * @param zkw zk reference
 * @param entityGroupName closed entityGroup to be deleted from zk
 * @throws org.apache.zookeeper.KeeperException if unexpected zookeeper exception
 * @throws org.apache.zookeeper.KeeperException.NoNodeException if node does not exist
 */
public static boolean deleteOfflineNode(ZooKeeperWatcher zkw,
  String entityGroupName) throws KeeperException,
  KeeperException.NoNodeException {
 return deleteNode(zkw, entityGroupName, EventType.M_ZK_ENTITYGROUP_OFFLINE);
}

代码示例来源：origin: alibaba/wasp

/**
 * @param zkw
 * @param pathOrEntityGroupName
 * @return Path to znode
 */
public static String getPath(final ZooKeeperWatcher zkw,
  final String pathOrEntityGroupName) {
 return pathOrEntityGroupName.startsWith("/") ? pathOrEntityGroupName : getNodeName(
   zkw, pathOrEntityGroupName);
}

代码示例来源：origin: alibaba/wasp

int versionid = ZKAssign.transitionNodeClosed(this.watcher,
  ENTITYGROUPINFO, SERVERNAME_A, -1);
assertNotSame(versionid, -1);
versionid = ZKAssign.getVersion(this.watcher, ENTITYGROUPINFO);
assertNotSame(-1, versionid);
versionid = ZKAssign.transitionNode(server.getZooKeeper(),
  ENTITYGROUPINFO, SERVERNAME_A, EventHandler.EventType.M_ZK_ENTITYGROUP_OFFLINE,
  EventHandler.EventType.FSERVER_ZK_ENTITYGROUP_OPENING, versionid);
assertNotSame(-1, versionid);
versionid = ZKAssign.transitionNodeOpened(this.watcher, ENTITYGROUPINFO,
  SERVERNAME_B, versionid);
assertNotSame(-1, versionid);
am.shutdown();
ZKAssign.deleteAllNodes(this.watcher);

代码示例来源：origin: alibaba/wasp

/**
 * Creates a znode with OPENED state.
 *
 * @param TEST_UTIL
 * @param entityGroup
 * @param serverName
 * @return
 * @throws java.io.IOException
 * @throws ZooKeeperConnectionException
 * @throws org.apache.zookeeper.KeeperException
 * @throws org.apache.zookeeper.KeeperException.NodeExistsException
 */
public static ZooKeeperWatcher createAndForceNodeToOpenedState(
  WaspTestingUtility TEST_UTIL, EntityGroup entityGroup,
  ServerName serverName) throws ZooKeeperConnectionException, IOException,
  KeeperException, NodeExistsException {
 ZooKeeperWatcher zkw = getZooKeeperWatcher(TEST_UTIL);
 ZKAssign.createNodeOffline(zkw, entityGroup.getEntityGroupInfo(),
   serverName);
 int version = ZKAssign.transitionNodeOpening(zkw,
   entityGroup.getEntityGroupInfo(), serverName);
 ZKAssign.transitionNodeOpened(zkw, entityGroup.getEntityGroupInfo(),
   serverName, version);
 return zkw;
}

代码示例来源：origin: alibaba/wasp

private void openEntityGroup(Server server, FServerServices rss, FTable ftd,
   EntityGroupInfo egi) throws IOException, NodeExistsException,
   KeeperException, DeserializationException {
  // Create it OFFLINE node, which is what Master set before sending OPEN RPC
  ZKAssign.createNodeOffline(server.getZooKeeper(), egi,
    server.getServerName());
  OpenEntityGroupHandler openHandler = new OpenEntityGroupHandler(server,
    rss, egi, ftd);
  openHandler.process();
  // This parse is not used?
  EntityGroupTransaction.parseFrom(ZKAssign.getData(server.getZooKeeper(),
    egi.getEncodedName()));
  // delete the node, which is what Master do after the entityGroup is opened
  ZKAssign.deleteNode(server.getZooKeeper(), egi.getEncodedName(),
    EventType.FSERVER_ZK_ENTITYGROUP_OPENED);
 }
}

代码示例来源：origin: alibaba/wasp

public static int transitionNodeOpening(ZooKeeperWatcher zkw,
  EntityGroupInfo entityGroup, ServerName serverName, final EventType beginState)
  throws KeeperException {
 return transitionNode(zkw, entityGroup, serverName, beginState,
   EventType.FSERVER_ZK_ENTITYGROUP_OPENING, -1);
}

代码示例来源：origin: alibaba/wasp

@Test
public void testFailedOpenEntityGroup() throws Exception {
 Server server = new MockServer(WTU);
 FServerServices rsServices = new MockFServerServices();
 // Create it OFFLINE, which is what it expects
 ZKAssign.createNodeOffline(server.getZooKeeper(), TEST_EGI,
   server.getServerName());
 // Create the handler
 OpenEntityGroupHandler handler = new OpenEntityGroupHandler(server,
   rsServices, TEST_EGI, TEST_FTD) {
  @Override
  EntityGroup openEntityGroup() {
   // Fake failure of opening a entityGroup due to an IOE, which is caught
   return null;
  }
 };
 handler.process();
 // Handler should have transitioned it to FAILED_OPEN
 EntityGroupTransaction rt = EntityGroupTransaction.parseFrom(ZKAssign
   .getData(server.getZooKeeper(), TEST_EGI.getEncodedName()));
 assertEquals(EventType.FSERVER_ZK_ENTITYGROUP_FAILED_OPEN,
   rt.getEventType());
}

代码示例来源：origin: alibaba/wasp

/**
 * Test if the entityGroup can be closed properly
 *
 * @throws java.io.IOException
 * @throws org.apache.zookeeper.KeeperException.NodeExistsException
 * @throws org.apache.zookeeper.KeeperException
 * @throws com.alibaba.wasp.DeserializationException
 */
@Test
public void testCloseEntityGroup() throws IOException, NodeExistsException,
  KeeperException, DeserializationException {
 final Server server = new MockServer(WTU);
 final MockFServerServices rss = new MockFServerServices();
 FTable htd = TEST_FTD;
 EntityGroupInfo egi = TEST_EGI;
 openEntityGroup(server, rss, htd, egi);
 int versionOfClosingNode = ZKAssign.createNodeClosing(
   server.getZooKeeper(), egi, server.getServerName());
 CloseEntityGroupHandler handler = new CloseEntityGroupHandler(server, rss,
   egi, false, true, versionOfClosingNode,
   EventType.M_FSERVER_CLOSE_ENTITYGROUP);
 handler.process();
 EntityGroupTransaction rt = EntityGroupTransaction.parseFrom(ZKAssign
   .getData(server.getZooKeeper(), egi.getEncodedName()));
 assertTrue(rt.getEventType()
   .equals(EventType.FSERVER_ZK_ENTITYGROUP_CLOSED));
}

代码示例来源：origin: alibaba/wasp

String path = ZKAssign.getNodeName(zkw, egi.getEncodedName());
Stat stats = zkw.getRecoverableZooKeeper().exists(path, false);
LOG.info("EPHEMERAL NODE BEFORE SERVER ABORT, path=" + path + ", stats="
  + stats);
EntityGroupTransaction rtd = EntityGroupTransaction.parseFrom(ZKAssign
  .getData(zkw, egi.getEncodedName()));

代码示例来源：origin: alibaba/wasp

HConstants.EMPTY_BYTE_ARRAY);
byte[] data = rt.toByteArray();
String node = getNodeName(zkw, entityGroup.getEncodedName());
zkw.sync(node);
int version = ZKUtil.checkExists(zkw, node);
  byte[] bytes = ZKAssign.getData(zkw, entityGroup.getEncodedName());
  rt = getEntityGroupTransition(bytes);
  if (rt.getEventType() != EventType.M_ZK_ENTITYGROUP_OFFLINE) {

代码示例来源：origin: alibaba/wasp

/**
 * Fakes the regionserver-side zk transitions of a region open.
 * @param w ZooKeeperWatcher to use.
 * @param sn Name of the regionserver doing the 'opening'
 * @param egInfo EntityGroup we're 'opening'.
 * @throws org.apache.zookeeper.KeeperException
 * @throws com.alibaba.wasp.DeserializationException
 */
static void fakeEntityGroupServerEntityGroupOpenInZK(FMaster master,  final ZooKeeperWatcher w,
  final ServerName sn, final EntityGroupInfo egInfo)
 throws KeeperException, DeserializationException, InterruptedException {
 // Wait till the we region is ready to be open in RIT.
 waitForEntityGroupPendingOpenInRIT(master.getAssignmentManager(), egInfo.getEncodedName());
 // Get current versionid else will fail on transition from OFFLINE to OPENING below
 int versionid = ZKAssign.getVersion(w, egInfo);
 assertNotSame(-1, versionid);
 // This uglyness below is what the openregionhandler on FSERVER side does.  I
 // looked at exposing the method over in openregionhandler but its just a
 // one liner and its deep over in another package so just repeat it below.
 versionid = ZKAssign.transitionNode(w, egInfo, sn,
   EventType.M_ZK_ENTITYGROUP_OFFLINE, EventType.FSERVER_ZK_ENTITYGROUP_OPENING, versionid);
 assertNotSame(-1, versionid);
 // Move znode from OPENING to OPENED as FSERVER does on successful open.
 versionid = ZKAssign.transitionNodeOpened(w, egInfo, sn, versionid);
 assertNotSame(-1, versionid);
 // We should be done now.  The master open handler will notice the
 // transition and remove this regions znode.
}

代码示例来源：origin: alibaba/wasp

LOG.debug(zkw.prefix("Deleting existing unassigned " + "node for "
  + entityGroupName + " that is in expected state " + expectedState));
String node = getNodeName(zkw, entityGroupName);
zkw.sync(node);
Stat stat = new Stat();
EntityGroupTransaction rt = getEntityGroupTransition(bytes);
EventType et = rt.getEventType();
if (!et.equals(expectedState)) {

代码示例来源：origin: alibaba/wasp

when(am.getZKTable()).thenReturn(new ZKTable(zkw));
Stat stat = new Stat();
String nodeName = ZKAssign.getNodeName(zkw, entityGroup
  .getEntityGroupInfo().getEncodedName());
ZKUtil.getDataAndWatch(zkw, nodeName, stat);
  stat.getVersion());
ZKAssign.transitionNode(zkw, entityGroup.getEntityGroupInfo(),
  server.getServerName(), EventType.FSERVER_ZK_ENTITYGROUP_OPENED,
  EventType.FSERVER_ZK_ENTITYGROUP_OPENED, stat.getVersion());

代码示例来源：origin: alibaba/wasp

String node = ZKAssign.getNodeName(watcher,
  entityGroupInfo.getEncodedName());
Stat stat = new Stat();
byte[] data = ZKAssign.getDataNoWatch(watcher, node, stat);
if (data == null) {
 LOG.warn("Data is null, node " + node + " no longer exists");

代码示例来源：origin: alibaba/wasp

ZKAssign.createNodeClosing(zkw, egi, new ServerName("any.old.server",
  1234, -1));
ZKAssign.deleteClosingNode(zkw, egi);

代码示例来源：origin: alibaba/wasp

@After
public void after() throws KeeperException {
 if (this.watcher != null) {
  // Clean up all znodes
  ZKAssign.deleteAllNodes(this.watcher);
  this.watcher.close();
 }
}

代码示例来源：origin: alibaba/wasp

/**
 * @param path
 * @return True if znode is in SPLIT or SPLITTING state.
 * @throws org.apache.zookeeper.KeeperException
 *           Can happen if the znode went away in meantime.
 * @throws com.alibaba.wasp.DeserializationException
 */
private boolean isSplitOrSplitting(final String path) throws KeeperException,
  DeserializationException {
 boolean result = false;
 // This may fail if the SPLIT or SPLITTING znode gets cleaned up before we
 // can get data from it.
 byte[] data = ZKAssign.getData(watcher, path);
 if (data == null)
  return false;
 EntityGroupTransaction rt = EntityGroupTransaction.parseFrom(data);
 switch (rt.getEventType()) {
 case FSERVER_ZK_ENTITYGROUP_SPLIT:
 case FSERVER_ZK_ENTITYGROUP_SPLITTING:
  result = true;
  break;
 default:
  break;
 }
 return result;
}

代码示例来源：origin: alibaba/wasp

return transitionNodeOpening(zkw, entityGroup, serverName,
  EventType.M_ZK_ENTITYGROUP_OFFLINE);

代码示例来源：origin: alibaba/wasp

final EntityGroup entityGroup) {
try {
 if (ZKAssign.transitionNodeClosed(server.getZooKeeper(), entityGroupInfo,
   server.getServerName(), expectedVersion) == FAILED) {
  LOG.warn("Completed the CLOSE of a entityGroup but when transitioning from "

内容来源于网络，如有侵权，请联系作者删除！

ZKAssign

相关文章

    查看更多

    热门标签

    更多
    JavaquerypythonNode开发语言requestUtil数据库Table后端算法LoggerMessageElementParser

    最新文章

    更多
    • 蜀ICP备13028337号-1 大数据知识库 https://www.saoniuhuo.com © All rights reserved
    • 本站内容来源互联网,如果侵犯您的权益请联系我们删除, 联系方式：448109455@qq.com