/** * Licensed to the Apache Software Foundation (ASF) under one * or more contributor license agreements. See the NOTICE file * distributed with this work for additional information * regarding copyright ownership. The ASF licenses this file * to you under the Apache License, Version 2.0 (the * "License"); you may not use this file except in compliance * with the License. You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, * software distributed under the License is distributed on an * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY * KIND, either express or implied. See the License for the * specific language governing permissions and limitations * under the License. */ package org.apache.hive.hcatalog.templeton.tool; import java.io.IOException; import java.util.ArrayList; import java.util.List; import org.slf4j.Logger; import org.slf4j.LoggerFactory; import org.apache.curator.framework.CuratorFramework; import org.apache.hadoop.conf.Configuration; import org.apache.zookeeper.CreateMode; import org.apache.zookeeper.KeeperException; import org.apache.zookeeper.ZooDefs.Ids; /* * The general idea here is to create * /created/1 * /created/2 * /created/3 .... * for each job submitted. The node number is generated by ZK (PERSISTENT_SEQUENTIAL) and the * payload is the JobId. Basically this keeps track of the order in which jobs were submitted, * and ZooKeeperCleanup uses this to purge old job info. * Since the /jobs/<id> node has a create/update timestamp * (http://zookeeper.apache.org/doc/trunk/zookeeperProgrammers.html#sc_zkStatStructure) this whole * thing can be removed. */ public class JobStateTracker { // The path to the tracking root private String job_trackingroot = null; // The zookeeper connection to use private CuratorFramework zk; // The id of the tracking node -- must be a SEQUENTIAL node private String trackingnode; // The id of the job this tracking node represents private String jobid; // The logger private static final Logger LOG = LoggerFactory.getLogger(JobStateTracker.class); /** * Constructor for a new node -- takes the jobid of an existing job * */ public JobStateTracker(String node, CuratorFramework zk, boolean nodeIsTracker, String job_trackingpath) { this.zk = zk; if (nodeIsTracker) { trackingnode = node; } else { jobid = node; } job_trackingroot = job_trackingpath; } /** * Create the parent znode for this job state. */ public void create() throws IOException { try { zk.create().creatingParentsIfNeeded().withMode(CreateMode.PERSISTENT) .withACL(Ids.OPEN_ACL_UNSAFE).forPath(job_trackingroot); } catch (KeeperException.NodeExistsException e) { //root must exist already } catch (Exception e) { throw new IOException("Unable to create parent nodes"); } try { trackingnode = zk.create().withMode(CreateMode.PERSISTENT_SEQUENTIAL) .withACL(Ids.OPEN_ACL_UNSAFE).forPath(makeTrackingZnode(), jobid.getBytes()); } catch (Exception e) { throw new IOException("Unable to create " + makeTrackingZnode()); } } public void delete() throws IOException { try { zk.delete().forPath(makeTrackingJobZnode(trackingnode)); } catch (Exception e) { // Might have been deleted already LOG.info("Couldn't delete " + makeTrackingJobZnode(trackingnode)); } } /** * Get the jobid for this tracking node * @throws IOException */ public String getJobID() throws IOException { try { return new String(zk.getData().forPath(makeTrackingJobZnode(trackingnode))); } catch (Exception e) { // It was deleted during the transaction throw new IOException("Node already deleted " + trackingnode, e); } } /** * Make a ZK path to a new tracking node */ public String makeTrackingZnode() { return job_trackingroot + "/"; } /** * Make a ZK path to an existing tracking node */ public String makeTrackingJobZnode(String nodename) { return job_trackingroot + "/" + nodename; } /* * Get the list of tracking jobs. These can be used to determine which jobs have * expired. */ public static List<String> getTrackingJobs(Configuration conf, CuratorFramework zk) throws IOException { ArrayList<String> jobs = new ArrayList<String>(); try { for (String myid : zk.getChildren().forPath( conf.get(TempletonStorage.STORAGE_ROOT) + ZooKeeperStorage.TRACKINGDIR)) { jobs.add(myid); } } catch (Exception e) { throw new IOException("Can't get tracking children", e); } return jobs; } }