Google Git
Sign in
apache / hadoop / a950c79f4fc6805bfe4515810190b408a0d38847 / . / src / mapred / org / apache / hadoop / mapred / TaskTracker.java
blob: d85484b6ac88c6705f39be0fef90522f1307d6e7 [file] [log] [blame]
/**
* 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.hadoop.mapred;
import java.io.ByteArrayOutputStream;
import java.io.DataInput;
import java.io.DataOutput;
import java.io.DataOutputStream;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.OutputStream;
import java.io.PrintStream;
import java.io.RandomAccessFile;
import java.net.InetSocketAddress;
import java.net.URI;
import java.net.URISyntaxException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.HashSet;
import java.util.Iterator;
import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Random;
import java.util.Set;
import java.util.TreeMap;
import java.util.LinkedHashMap;
import java.util.Vector;
import java.util.concurrent.BlockingQueue;
import java.util.concurrent.LinkedBlockingQueue;
import java.util.regex.Pattern;
import javax.servlet.ServletContext;
import javax.servlet.ServletException;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.filecache.DistributedCache;
import org.apache.hadoop.fs.DF;
import org.apache.hadoop.fs.FSDataInputStream;
import org.apache.hadoop.fs.FSError;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.FileUtil;
import org.apache.hadoop.fs.LocalDirAllocator;
import org.apache.hadoop.fs.LocalFileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.io.WritableUtils;
import org.apache.hadoop.ipc.RPC;
import org.apache.hadoop.ipc.RemoteException;
import org.apache.hadoop.ipc.Server;
import org.apache.hadoop.mapred.JobClient.TaskStatusFilter;
import org.apache.hadoop.mapred.TaskStatus.Phase;
import org.apache.hadoop.mapred.pipes.Submitter;
import org.apache.hadoop.metrics.MetricsContext;
import org.apache.hadoop.metrics.MetricsException;
import org.apache.hadoop.metrics.MetricsRecord;
import org.apache.hadoop.metrics.MetricsUtil;
import org.apache.hadoop.metrics.Updater;
import org.apache.hadoop.metrics.jvm.JvmMetrics;
import org.apache.hadoop.net.DNS;
import org.apache.hadoop.net.NetUtils;
import org.apache.hadoop.util.DiskChecker;
import org.apache.hadoop.util.ProcfsBasedProcessTree;
import org.apache.hadoop.util.ReflectionUtils;
import org.apache.hadoop.util.RunJar;
import org.apache.hadoop.util.StringUtils;
import org.apache.hadoop.util.VersionInfo;
import org.apache.hadoop.util.DiskChecker.DiskErrorException;
import org.apache.hadoop.util.Shell.ShellCommandExecutor;
import org.apache.log4j.LogManager;
/*******************************************************
* TaskTracker is a process that starts and tracks MR Tasks
* in a networked environment. It contacts the JobTracker
* for Task assignments and reporting results.
*
*******************************************************/
public class TaskTracker
implements MRConstants, TaskUmbilicalProtocol, Runnable {
static final long WAIT_FOR_DONE = 3 * 1000;
private int httpPort;
static enum State {NORMAL, STALE, INTERRUPTED, DENIED}
public static final Log LOG =
LogFactory.getLog(TaskTracker.class);
public static final String MR_CLIENTTRACE_FORMAT =
"src: %s" + // src IP
", dest: %s" + // dst IP
", bytes: %s" + // byte count
", op: %s" + // operation
", cliID: %s"; // task id
public static final Log ClientTraceLog =
LogFactory.getLog(TaskTracker.class.getName() + ".clienttrace");
volatile boolean running = true;
private LocalDirAllocator localDirAllocator;
String taskTrackerName;
String localHostname;
InetSocketAddress jobTrackAddr;
InetSocketAddress taskReportAddress;
Server taskReportServer = null;
InterTrackerProtocol jobClient;
// last heartbeat response recieved
short heartbeatResponseId = -1;
/*
* This is the last 'status' report sent by this tracker to the JobTracker.
*
* If the rpc call succeeds, this 'status' is cleared-out by this tracker;
* indicating that a 'fresh' status report be generated; in the event the
* rpc calls fails for whatever reason, the previous status report is sent
* again.
*/
TaskTrackerStatus status = null;
// The system-directory on HDFS where job files are stored
Path systemDirectory = null;
// The filesystem where job files are stored
FileSystem systemFS = null;
StatusHttpServer server = null;
volatile boolean shuttingDown = false;
Map<TaskAttemptID, TaskInProgress> tasks = new HashMap<TaskAttemptID, TaskInProgress>();
/**
* Map from taskId -> TaskInProgress.
*/
Map<TaskAttemptID, TaskInProgress> runningTasks = null;
Map<JobID, RunningJob> runningJobs = null;
volatile int mapTotal = 0;
volatile int reduceTotal = 0;
boolean justStarted = true;
// Mark reduce tasks that are shuffling to rollback their events index
Set<TaskAttemptID> shouldReset = new HashSet<TaskAttemptID>();
//dir -> DF
Map<String, DF> localDirsDf = new HashMap<String, DF>();
long minSpaceStart = 0;
//must have this much space free to start new tasks
boolean acceptNewTasks = true;
long minSpaceKill = 0;
//if we run under this limit, kill one task
//and make sure we never receive any new jobs
//until all the old tasks have been cleaned up.
//this is if a machine is so full it's only good
//for serving map output to the other nodes
static Random r = new Random();
private static final String SUBDIR = "taskTracker";
private static final String CACHEDIR = "archive";
private static final String JOBCACHE = "jobcache";
private static final String PID = "pid";
private static final String OUTPUT = "output";
private JobConf originalConf;
private JobConf fConf;
private int maxCurrentMapTasks;
private int maxCurrentReduceTasks;
private int failures;
private int finishedCount[] = new int[1];
private MapEventsFetcherThread mapEventsFetcher;
int workerThreads;
private CleanupQueue directoryCleanupThread;
volatile JvmManager jvmManager;
private TaskMemoryManagerThread taskMemoryManager;
private boolean taskMemoryManagerEnabled = false;
private long maxVirtualMemoryForTasks
= JobConf.DISABLED_VIRTUAL_MEMORY_LIMIT;
/**
* the minimum interval between jobtracker polls
*/
private volatile int heartbeatInterval = HEARTBEAT_INTERVAL_MIN;
/**
* Number of maptask completion events locations to poll for at one time
*/
private int probe_sample_size = 500;
private IndexCache indexCache;
/*
* A list of commitTaskActions for whom commit response has been received
*/
private List<TaskAttemptID> commitResponses =
Collections.synchronizedList(new ArrayList<TaskAttemptID>());
private ShuffleServerMetrics shuffleServerMetrics;
/** This class contains the methods that should be used for metrics-reporting
* the specific metrics for shuffle. The TaskTracker is actually a server for
* the shuffle and hence the name ShuffleServerMetrics.
*/
private class ShuffleServerMetrics implements Updater {
private MetricsRecord shuffleMetricsRecord = null;
private int serverHandlerBusy = 0;
private long outputBytes = 0;
private int failedOutputs = 0;
private int successOutputs = 0;
ShuffleServerMetrics(JobConf conf) {
MetricsContext context = MetricsUtil.getContext("mapred");
shuffleMetricsRecord =
MetricsUtil.createRecord(context, "shuffleOutput");
this.shuffleMetricsRecord.setTag("sessionId", conf.getSessionId());
context.registerUpdater(this);
}
synchronized void serverHandlerBusy() {
++serverHandlerBusy;
}
synchronized void serverHandlerFree() {
--serverHandlerBusy;
}
synchronized void outputBytes(long bytes) {
outputBytes += bytes;
}
synchronized void failedOutput() {
++failedOutputs;
}
synchronized void successOutput() {
++successOutputs;
}
public void doUpdates(MetricsContext unused) {
synchronized (this) {
if (workerThreads != 0) {
shuffleMetricsRecord.setMetric("shuffle_handler_busy_percent",
100*((float)serverHandlerBusy/workerThreads));
} else {
shuffleMetricsRecord.setMetric("shuffle_handler_busy_percent", 0);
}
shuffleMetricsRecord.incrMetric("shuffle_output_bytes",
outputBytes);
shuffleMetricsRecord.incrMetric("shuffle_failed_outputs",
failedOutputs);
shuffleMetricsRecord.incrMetric("shuffle_success_outputs",
successOutputs);
outputBytes = 0;
failedOutputs = 0;
successOutputs = 0;
}
shuffleMetricsRecord.update();
}
}
private TaskTrackerInstrumentation myInstrumentation = null;
public TaskTrackerInstrumentation getTaskTrackerInstrumentation() {
return myInstrumentation;
}
/**
* A list of tips that should be cleaned up.
*/
private BlockingQueue<TaskTrackerAction> tasksToCleanup =
new LinkedBlockingQueue<TaskTrackerAction>();
/**
* A daemon-thread that pulls tips off the list of things to cleanup.
*/
private Thread taskCleanupThread =
new Thread(new Runnable() {
public void run() {
while (true) {
try {
TaskTrackerAction action = tasksToCleanup.take();
if (action instanceof KillJobAction) {
purgeJob((KillJobAction) action);
} else if (action instanceof KillTaskAction) {
TaskInProgress tip;
KillTaskAction killAction = (KillTaskAction) action;
synchronized (TaskTracker.this) {
tip = tasks.get(killAction.getTaskID());
}
LOG.info("Received KillTaskAction for task: " +
killAction.getTaskID());
purgeTask(tip, false);
} else {
LOG.error("Non-delete action given to cleanup thread: "
+ action);
}
} catch (Throwable except) {
LOG.warn(StringUtils.stringifyException(except));
}
}
}
}, "taskCleanup");
private RunningJob addTaskToJob(JobID jobId,
TaskInProgress tip) {
synchronized (runningJobs) {
RunningJob rJob = null;
if (!runningJobs.containsKey(jobId)) {
rJob = new RunningJob(jobId);
rJob.localized = false;
rJob.tasks = new HashSet<TaskInProgress>();
runningJobs.put(jobId, rJob);
} else {
rJob = runningJobs.get(jobId);
}
synchronized (rJob) {
rJob.tasks.add(tip);
}
runningJobs.notify(); //notify the fetcher thread
return rJob;
}
}
private void removeTaskFromJob(JobID jobId, TaskInProgress tip) {
synchronized (runningJobs) {
RunningJob rjob = runningJobs.get(jobId);
if (rjob == null) {
LOG.warn("Unknown job " + jobId + " being deleted.");
} else {
synchronized (rjob) {
rjob.tasks.remove(tip);
}
}
}
}
static String getCacheSubdir() {
return TaskTracker.SUBDIR + Path.SEPARATOR + TaskTracker.CACHEDIR;
}
static String getJobCacheSubdir() {
return TaskTracker.SUBDIR + Path.SEPARATOR + TaskTracker.JOBCACHE;
}
static String getLocalJobDir(String jobid) {
return getJobCacheSubdir() + Path.SEPARATOR + jobid;
}
static String getLocalTaskDir(String jobid, String taskid) {
return getLocalTaskDir(jobid, taskid, false) ;
}
static String getIntermediateOutputDir(String jobid, String taskid) {
return getLocalTaskDir(jobid, taskid)
+ Path.SEPARATOR + TaskTracker.OUTPUT ;
}
static String getLocalTaskDir(String jobid,
String taskid,
boolean isCleanupAttempt) {
String taskDir = getLocalJobDir(jobid) + Path.SEPARATOR + taskid;
if (isCleanupAttempt) {
taskDir = taskDir + ".cleanup";
}
return taskDir;
}
static String getPidFile(String jobid,
String taskid,
boolean isCleanup) {
return getLocalTaskDir(jobid, taskid, isCleanup)
+ Path.SEPARATOR + PID;
}
public long getProtocolVersion(String protocol,
long clientVersion) throws IOException {
if (protocol.equals(TaskUmbilicalProtocol.class.getName())) {
return TaskUmbilicalProtocol.versionID;
} else {
throw new IOException("Unknown protocol for task tracker: " +
protocol);
}
}
/**
* Do the real constructor work here. It's in a separate method
* so we can call it again and "recycle" the object after calling
* close().
*/
synchronized void initialize() throws IOException {
// use configured nameserver & interface to get local hostname
this.fConf = new JobConf(originalConf);
if (fConf.get("slave.host.name") != null) {
this.localHostname = fConf.get("slave.host.name");
}
if (localHostname == null) {
this.localHostname =
DNS.getDefaultHost
(fConf.get("mapred.tasktracker.dns.interface","default"),
fConf.get("mapred.tasktracker.dns.nameserver","default"));
}
//check local disk
checkLocalDirs(this.fConf.getLocalDirs());
fConf.deleteLocalFiles(SUBDIR);
// Clear out state tables
this.tasks.clear();
this.runningTasks = new LinkedHashMap<TaskAttemptID, TaskInProgress>();
this.runningJobs = new TreeMap<JobID, RunningJob>();
this.mapTotal = 0;
this.reduceTotal = 0;
this.acceptNewTasks = true;
this.status = null;
this.minSpaceStart = this.fConf.getLong("mapred.local.dir.minspacestart", 0L);
this.minSpaceKill = this.fConf.getLong("mapred.local.dir.minspacekill", 0L);
//tweak the probe sample size (make it a function of numCopiers)
probe_sample_size = this.fConf.getInt("mapred.tasktracker.events.batchsize", 500);
Class<? extends TaskTrackerInstrumentation> metricsInst = getInstrumentationClass(fConf);
try {
java.lang.reflect.Constructor<? extends TaskTrackerInstrumentation> c =
metricsInst.getConstructor(new Class[] {TaskTracker.class} );
this.myInstrumentation = c.newInstance(this);
} catch(Exception e) {
//Reflection can throw lots of exceptions -- handle them all by
//falling back on the default.
LOG.error("failed to initialize taskTracker metrics", e);
this.myInstrumentation = new TaskTrackerMetricsInst(this);
}
// bind address
String address =
NetUtils.getServerAddress(fConf,
"mapred.task.tracker.report.bindAddress",
"mapred.task.tracker.report.port",
"mapred.task.tracker.report.address");
InetSocketAddress socAddr = NetUtils.createSocketAddr(address);
String bindAddress = socAddr.getHostName();
int tmpPort = socAddr.getPort();
this.jvmManager = new JvmManager(this);
// RPC initialization
int max = maxCurrentMapTasks > maxCurrentReduceTasks ?
maxCurrentMapTasks : maxCurrentReduceTasks;
//set the num handlers to max*2 since canCommit may wait for the duration
//of a heartbeat RPC
this.taskReportServer =
RPC.getServer(this, bindAddress, tmpPort, 2 * max, false, this.fConf);
this.taskReportServer.start();
// get the assigned address
this.taskReportAddress = taskReportServer.getListenerAddress();
this.fConf.set("mapred.task.tracker.report.address",
taskReportAddress.getHostName() + ":" + taskReportAddress.getPort());
LOG.info("TaskTracker up at: " + this.taskReportAddress);
this.taskTrackerName = "tracker_" + localHostname + ":" + taskReportAddress;
LOG.info("Starting tracker " + taskTrackerName);
// Clear out temporary files that might be lying around
DistributedCache.purgeCache(this.fConf);
cleanupStorage();
this.justStarted = true;
this.jobClient = (InterTrackerProtocol)
RPC.waitForProxy(InterTrackerProtocol.class,
InterTrackerProtocol.versionID,
jobTrackAddr, this.fConf);
this.running = true;
// start the thread that will fetch map task completion events
this.mapEventsFetcher = new MapEventsFetcherThread();
mapEventsFetcher.setDaemon(true);
mapEventsFetcher.setName(
"Map-events fetcher for all reduce tasks " + "on " +
taskTrackerName);
mapEventsFetcher.start();
maxVirtualMemoryForTasks = fConf.
getLong("mapred.tasktracker.tasks.maxmemory",
JobConf.DISABLED_VIRTUAL_MEMORY_LIMIT);
this.indexCache = new IndexCache(this.fConf);
// start the taskMemoryManager thread only if enabled
setTaskMemoryManagerEnabledFlag();
if (isTaskMemoryManagerEnabled()) {
taskMemoryManager = new TaskMemoryManagerThread(this);
taskMemoryManager.setDaemon(true);
taskMemoryManager.start();
}
mapLauncher = new TaskLauncher(maxCurrentMapTasks);
reduceLauncher = new TaskLauncher(maxCurrentReduceTasks);
mapLauncher.start();
reduceLauncher.start();
}
public static Class<? extends TaskTrackerInstrumentation> getInstrumentationClass(Configuration conf) {
return conf.getClass("mapred.tasktracker.instrumentation",
TaskTrackerMetricsInst.class, TaskTrackerInstrumentation.class);
}
public static void setInstrumentationClass(Configuration conf, Class<? extends TaskTrackerInstrumentation> t) {
conf.setClass("mapred.tasktracker.instrumentation",
t, TaskTrackerInstrumentation.class);
}
/**
* Removes all contents of temporary storage. Called upon
* startup, to remove any leftovers from previous run.
*/
public void cleanupStorage() throws IOException {
this.fConf.deleteLocalFiles();
}
// Object on wait which MapEventsFetcherThread is going to wait.
private Object waitingOn = new Object();
private class MapEventsFetcherThread extends Thread {
private List <FetchStatus> reducesInShuffle() {
List <FetchStatus> fList = new ArrayList<FetchStatus>();
for (Map.Entry <JobID, RunningJob> item : runningJobs.entrySet()) {
RunningJob rjob = item.getValue();
JobID jobId = item.getKey();
FetchStatus f;
synchronized (rjob) {
f = rjob.getFetchStatus();
for (TaskInProgress tip : rjob.tasks) {
Task task = tip.getTask();
if (!task.isMapTask()) {
if (((ReduceTask)task).getPhase() ==
TaskStatus.Phase.SHUFFLE) {
if (rjob.getFetchStatus() == null) {
//this is a new job; we start fetching its map events
f = new FetchStatus(jobId,
((ReduceTask)task).getNumMaps());
rjob.setFetchStatus(f);
}
f = rjob.getFetchStatus();
fList.add(f);
break; //no need to check any more tasks belonging to this
}
}
}
}
}
//at this point, we have information about for which of
//the running jobs do we need to query the jobtracker for map
//outputs (actually map events).
return fList;
}
@Override
public void run() {
LOG.info("Starting thread: " + getName());
while (running) {
try {
List <FetchStatus> fList = null;
synchronized (runningJobs) {
while (((fList = reducesInShuffle()).size()) == 0) {
try {
runningJobs.wait();
} catch (InterruptedException e) {
LOG.info("Shutting down: " + getName());
return;
}
}
}
// now fetch all the map task events for all the reduce tasks
// possibly belonging to different jobs
boolean fetchAgain = false; //flag signifying whether we want to fetch
//immediately again.
for (FetchStatus f : fList) {
long currentTime = System.currentTimeMillis();
try {
//the method below will return true when we have not
//fetched all available events yet
if (f.fetchMapCompletionEvents(currentTime)) {
fetchAgain = true;
}
} catch (Exception e) {
LOG.warn(
"Ignoring exception that fetch for map completion" +
" events threw for " + f.jobId + " threw: " +
StringUtils.stringifyException(e));
}
if (!running) {
break;
}
}
synchronized (waitingOn) {
try {
int waitTime;
if (!fetchAgain) {
waitingOn.wait(heartbeatInterval);
}
} catch (InterruptedException ie) {
LOG.info("Shutting down: " + getName());
return;
}
}
} catch (Exception e) {
LOG.info("Ignoring exception " + e.getMessage());
}
}
}
}
private class FetchStatus {
/** The next event ID that we will start querying the JobTracker from*/
private IntWritable fromEventId;
/** This is the cache of map events for a given job */
private List<TaskCompletionEvent> allMapEvents;
/** What jobid this fetchstatus object is for*/
private JobID jobId;
private long lastFetchTime;
private boolean fetchAgain;
public FetchStatus(JobID jobId, int numMaps) {
this.fromEventId = new IntWritable(0);
this.jobId = jobId;
this.allMapEvents = new ArrayList<TaskCompletionEvent>(numMaps);
}
/**
* Check if the number of events that are obtained are more than required.
* If yes then purge the extra ones.
*/
public void purgeMapEvents(int lastKnownIndex) {
// Note that the sync is first on fromEventId and then on allMapEvents
synchronized (fromEventId) {
synchronized (allMapEvents) {
int index = 0;
if (allMapEvents.size() > lastKnownIndex) {
fromEventId.set(lastKnownIndex);
allMapEvents = allMapEvents.subList(0, lastKnownIndex);
}
}
}
}
public TaskCompletionEvent[] getMapEvents(int fromId, int max) {
TaskCompletionEvent[] mapEvents =
TaskCompletionEvent.EMPTY_ARRAY;
boolean notifyFetcher = false;
synchronized (allMapEvents) {
if (allMapEvents.size() > fromId) {
int actualMax = Math.min(max, (allMapEvents.size() - fromId));
List <TaskCompletionEvent> eventSublist =
allMapEvents.subList(fromId, actualMax + fromId);
mapEvents = eventSublist.toArray(mapEvents);
} else {
// Notify Fetcher thread.
notifyFetcher = true;
}
}
if (notifyFetcher) {
synchronized (waitingOn) {
waitingOn.notify();
}
}
return mapEvents;
}
public boolean fetchMapCompletionEvents(long currTime) throws IOException {
if (!fetchAgain && (currTime - lastFetchTime) < heartbeatInterval) {
return false;
}
int currFromEventId = 0;
synchronized (fromEventId) {
currFromEventId = fromEventId.get();
List <TaskCompletionEvent> recentMapEvents =
queryJobTracker(fromEventId, jobId, jobClient);
synchronized (allMapEvents) {
allMapEvents.addAll(recentMapEvents);
}
lastFetchTime = currTime;
if (fromEventId.get() - currFromEventId >= probe_sample_size) {
//return true when we have fetched the full payload, indicating
//that we should fetch again immediately (there might be more to
//fetch
fetchAgain = true;
return true;
}
}
fetchAgain = false;
return false;
}
}
private LocalDirAllocator lDirAlloc =
new LocalDirAllocator("mapred.local.dir");
// intialize the job directory
private void localizeJob(TaskInProgress tip) throws IOException {
Path localJarFile = null;
Task t = tip.getTask();
JobID jobId = t.getJobID();
Path jobFile = new Path(t.getJobFile());
// Get sizes of JobFile and JarFile
// sizes are -1 if they are not present.
FileStatus status = null;
long jobFileSize = -1;
try {
status = systemFS.getFileStatus(jobFile);
jobFileSize = status.getLen();
} catch(FileNotFoundException fe) {
jobFileSize = -1;
}
Path localJobFile = lDirAlloc.getLocalPathForWrite(
getLocalJobDir(jobId.toString())
+ Path.SEPARATOR + "job.xml",
jobFileSize, fConf);
RunningJob rjob = addTaskToJob(jobId, tip);
synchronized (rjob) {
if (!rjob.localized) {
FileSystem localFs = FileSystem.getLocal(fConf);
// this will happen on a partial execution of localizeJob.
// Sometimes the job.xml gets copied but copying job.jar
// might throw out an exception
// we should clean up and then try again
Path jobDir = localJobFile.getParent();
if (localFs.exists(jobDir)){
localFs.delete(jobDir, true);
boolean b = localFs.mkdirs(jobDir);
if (!b)
throw new IOException("Not able to create job directory "
+ jobDir.toString());
}
systemFS.copyToLocalFile(jobFile, localJobFile);
JobConf localJobConf = new JobConf(localJobFile);
// create the 'work' directory
// job-specific shared directory for use as scratch space
Path workDir = lDirAlloc.getLocalPathForWrite(
(getLocalJobDir(jobId.toString())
+ Path.SEPARATOR + "work"), fConf);
if (!localFs.mkdirs(workDir)) {
throw new IOException("Mkdirs failed to create "
+ workDir.toString());
}
System.setProperty("job.local.dir", workDir.toString());
localJobConf.set("job.local.dir", workDir.toString());
// copy Jar file to the local FS and unjar it.
String jarFile = localJobConf.getJar();
long jarFileSize = -1;
if (jarFile != null) {
Path jarFilePath = new Path(jarFile);
try {
status = systemFS.getFileStatus(jarFilePath);
jarFileSize = status.getLen();
} catch(FileNotFoundException fe) {
jarFileSize = -1;
}
// Here we check for and we check five times the size of jarFileSize
// to accommodate for unjarring the jar file in work directory
localJarFile = new Path(lDirAlloc.getLocalPathForWrite(
getLocalJobDir(jobId.toString())
+ Path.SEPARATOR + "jars",
5 * jarFileSize, fConf), "job.jar");
if (!localFs.mkdirs(localJarFile.getParent())) {
throw new IOException("Mkdirs failed to create jars directory ");
}
systemFS.copyToLocalFile(jarFilePath, localJarFile);
localJobConf.setJar(localJarFile.toString());
OutputStream out = localFs.create(localJobFile);
try {
localJobConf.writeXml(out);
} finally {
out.close();
}
// also unjar the job.jar files
RunJar.unJar(new File(localJarFile.toString()),
new File(localJarFile.getParent().toString()));
}
rjob.keepJobFiles = ((localJobConf.getKeepTaskFilesPattern() != null) ||
localJobConf.getKeepFailedTaskFiles());
rjob.localized = true;
rjob.jobConf = localJobConf;
}
}
launchTaskForJob(tip, new JobConf(rjob.jobConf));
}
private void launchTaskForJob(TaskInProgress tip, JobConf jobConf) throws IOException{
synchronized (tip) {
tip.setJobConf(jobConf);
tip.launchTask();
}
}
public synchronized void shutdown() throws IOException {
shuttingDown = true;
close();
if (this.server != null) {
try {
LOG.info("Shutting down StatusHttpServer");
this.server.stop();
} catch (InterruptedException ex) {
ex.printStackTrace();
}
}
}
/**
* Close down the TaskTracker and all its components. We must also shutdown
* any running tasks or threads, and cleanup disk space. A new TaskTracker
* within the same process space might be restarted, so everything must be
* clean.
*/
public synchronized void close() throws IOException {
//
// Kill running tasks. Do this in a 2nd vector, called 'tasksToClose',
// because calling jobHasFinished() may result in an edit to 'tasks'.
//
TreeMap<TaskAttemptID, TaskInProgress> tasksToClose =
new TreeMap<TaskAttemptID, TaskInProgress>();
tasksToClose.putAll(tasks);
for (TaskInProgress tip : tasksToClose.values()) {
tip.jobHasFinished(false);
}
this.running = false;
// Clear local storage
cleanupStorage();
// Shutdown the fetcher thread
this.mapEventsFetcher.interrupt();
//stop the launchers
this.mapLauncher.interrupt();
this.reduceLauncher.interrupt();
jvmManager.stop();
// shutdown RPC connections
RPC.stopProxy(jobClient);
// wait for the fetcher thread to exit
for (boolean done = false; !done; ) {
try {
this.mapEventsFetcher.join();
done = true;
} catch (InterruptedException e) {
}
}
if (taskReportServer != null) {
taskReportServer.stop();
taskReportServer = null;
}
}
/**
* Start with the local machine name, and the default JobTracker
*/
public TaskTracker(JobConf conf) throws IOException {
originalConf = conf;
maxCurrentMapTasks = conf.getInt(
"mapred.tasktracker.map.tasks.maximum", 2);
maxCurrentReduceTasks = conf.getInt(
"mapred.tasktracker.reduce.tasks.maximum", 2);
this.jobTrackAddr = JobTracker.getAddress(conf);
String infoAddr =
NetUtils.getServerAddress(conf,
"tasktracker.http.bindAddress",
"tasktracker.http.port",
"mapred.task.tracker.http.address");
InetSocketAddress infoSocAddr = NetUtils.createSocketAddr(infoAddr);
String httpBindAddress = infoSocAddr.getHostName();
int httpPort = infoSocAddr.getPort();
this.server = new StatusHttpServer("task", httpBindAddress, httpPort,
httpPort == 0, conf);
workerThreads = conf.getInt("tasktracker.http.threads", 40);
this.shuffleServerMetrics = new ShuffleServerMetrics(conf);
server.setThreads(1, workerThreads);
// let the jsp pages get to the task tracker, config, and other relevant
// objects
FileSystem local = FileSystem.getLocal(conf);
this.localDirAllocator = new LocalDirAllocator("mapred.local.dir");
server.setAttribute("task.tracker", this);
server.setAttribute("local.file.system", local);
server.setAttribute("conf", conf);
server.setAttribute("log", LOG);
server.setAttribute("localDirAllocator", localDirAllocator);
server.setAttribute("shuffleServerMetrics", shuffleServerMetrics);
server.addInternalServlet("mapOutput", "/mapOutput", MapOutputServlet.class);
server.addInternalServlet("taskLog", "/tasklog", TaskLogServlet.class);
server.start();
this.httpPort = server.getPort();
initialize();
}
private void startCleanupThreads() throws IOException {
taskCleanupThread.setDaemon(true);
taskCleanupThread.start();
directoryCleanupThread = new CleanupQueue();
}
/**
* The connection to the JobTracker, used by the TaskRunner
* for locating remote files.
*/
public InterTrackerProtocol getJobClient() {
return jobClient;
}
/** Return the port at which the tasktracker bound to */
public synchronized InetSocketAddress getTaskTrackerReportAddress() {
return taskReportAddress;
}
/** Queries the job tracker for a set of outputs ready to be copied
* @param fromEventId the first event ID we want to start from, this is
* modified by the call to this method
* @param jobClient the job tracker
* @return a set of locations to copy outputs from
* @throws IOException
*/
private List<TaskCompletionEvent> queryJobTracker(IntWritable fromEventId,
JobID jobId,
InterTrackerProtocol jobClient)
throws IOException {
TaskCompletionEvent t[] = jobClient.getTaskCompletionEvents(
jobId,
fromEventId.get(),
probe_sample_size);
//we are interested in map task completion events only. So store
//only those
List <TaskCompletionEvent> recentMapEvents =
new ArrayList<TaskCompletionEvent>();
for (int i = 0; i < t.length; i++) {
if (t[i].isMap) {
recentMapEvents.add(t[i]);
}
}
fromEventId.set(fromEventId.get() + t.length);
return recentMapEvents;
}
/**
* Main service loop. Will stay in this loop forever.
*/
State offerService() throws Exception {
long lastHeartbeat = 0;
while (running && !shuttingDown) {
try {
long now = System.currentTimeMillis();
long waitTime = heartbeatInterval - (now - lastHeartbeat);
if (waitTime > 0) {
// sleeps for the wait time, wakes up if a task is finished.
synchronized(finishedCount) {
if (finishedCount[0] == 0) {
finishedCount.wait(waitTime);
}
finishedCount[0] = 0;
}
}
// If the TaskTracker is just starting up:
// 1. Verify the buildVersion
// 2. Get the system directory & filesystem
if(justStarted){
String jobTrackerBV = jobClient.getBuildVersion();
if(!VersionInfo.getBuildVersion().equals(jobTrackerBV)) {
String msg = "Shutting down. Incompatible buildVersion." +
"\nJobTracker's: " + jobTrackerBV +
"\nTaskTracker's: "+ VersionInfo.getBuildVersion();
LOG.error(msg);
try {
jobClient.reportTaskTrackerError(taskTrackerName, null, msg);
} catch(Exception e ) {
LOG.info("Problem reporting to jobtracker: " + e);
}
return State.DENIED;
}
String dir = jobClient.getSystemDir();
if (dir == null) {
throw new IOException("Failed to get system directory");
}
systemDirectory = new Path(dir);
systemFS = systemDirectory.getFileSystem(fConf);
}
// Send the heartbeat and process the jobtracker's directives
HeartbeatResponse heartbeatResponse = transmitHeartBeat(now);
// Note the time when the heartbeat returned, use this to decide when to send the
// next heartbeat
lastHeartbeat = System.currentTimeMillis();
// Check if the map-event list needs purging
if (heartbeatResponse.getLastKnownIndex() != null) {
synchronized (this) {
// purge the local map events list
for (Map.Entry<JobID, Integer> entry
: heartbeatResponse.getLastKnownIndex().entrySet()) {
RunningJob rjob;
synchronized (runningJobs) {
rjob = runningJobs.get(entry.getKey());
if (rjob != null) {
synchronized (rjob) {
FetchStatus f = rjob.getFetchStatus();
if (f != null) {
f.purgeMapEvents(entry.getValue());
}
}
}
}
}
// Mark the reducers in shuffle for rollback
synchronized (shouldReset) {
for (Map.Entry<TaskAttemptID, TaskInProgress> entry
: runningTasks.entrySet()) {
if (entry.getValue().getStatus().getPhase() == Phase.SHUFFLE) {
this.shouldReset.add(entry.getKey());
}
}
}
}
}
TaskTrackerAction[] actions = heartbeatResponse.getActions();
if(LOG.isDebugEnabled()) {
LOG.debug("Got heartbeatResponse from JobTracker with responseId: " +
heartbeatResponse.getResponseId() + " and " +
((actions != null) ? actions.length : 0) + " actions");
}
if (reinitTaskTracker(actions)) {
return State.STALE;
}
// resetting heartbeat interval from the response.
heartbeatInterval = heartbeatResponse.getHeartbeatInterval();
justStarted = false;
if (actions != null){
for(TaskTrackerAction action: actions) {
if (action instanceof LaunchTaskAction) {
addToTaskQueue((LaunchTaskAction)action);
} else if (action instanceof CommitTaskAction) {
CommitTaskAction commitAction = (CommitTaskAction)action;
if (!commitResponses.contains(commitAction.getTaskID())) {
LOG.info("Received commit task action for " +
commitAction.getTaskID());
commitResponses.add(commitAction.getTaskID());
}
} else {
tasksToCleanup.put(action);
}
}
}
markUnresponsiveTasks();
killOverflowingTasks();
//we've cleaned up, resume normal operation
if (!acceptNewTasks && isIdle()) {
acceptNewTasks=true;
}
} catch (InterruptedException ie) {
LOG.info("Interrupted. Closing down.");
return State.INTERRUPTED;
} catch (DiskErrorException de) {
String msg = "Exiting task tracker for disk error:\n" +
StringUtils.stringifyException(de);
LOG.error(msg);
synchronized (this) {
jobClient.reportTaskTrackerError(taskTrackerName,
"DiskErrorException", msg);
}
return State.STALE;
} catch (RemoteException re) {
String reClass = re.getClassName();
if (DisallowedTaskTrackerException.class.getName().equals(reClass)) {
LOG.info("Tasktracker disallowed by JobTracker.");
return State.DENIED;
}
} catch (Exception except) {
String msg = "Caught exception: " +
StringUtils.stringifyException(except);
LOG.error(msg);
}
}
return State.NORMAL;
}
private long previousUpdate = 0;
/**
* Build and transmit the heart beat to the JobTracker
* @param now current time
* @return false if the tracker was unknown
* @throws IOException
*/
private HeartbeatResponse transmitHeartBeat(long now) throws IOException {
// Send Counters in the status once every COUNTER_UPDATE_INTERVAL
boolean sendCounters;
if (now > (previousUpdate + COUNTER_UPDATE_INTERVAL)) {
sendCounters = true;
previousUpdate = now;
}
else {
sendCounters = false;
}
//
// Check if the last heartbeat got through...
// if so then build the heartbeat information for the JobTracker;
// else resend the previous status information.
//
if (status == null) {
synchronized (this) {
status = new TaskTrackerStatus(taskTrackerName, localHostname,
httpPort,
cloneAndResetRunningTaskStatuses(
sendCounters),
failures,
maxCurrentMapTasks,
maxCurrentReduceTasks);
}
} else {
LOG.info("Resending 'status' to '" + jobTrackAddr.getHostName() +
"' with reponseId '" + heartbeatResponseId);
}
//
// Check if we should ask for a new Task
//
boolean askForNewTask;
long localMinSpaceStart;
synchronized (this) {
askForNewTask = (status.countMapTasks() < maxCurrentMapTasks ||
status.countReduceTasks() < maxCurrentReduceTasks) &&
acceptNewTasks;
localMinSpaceStart = minSpaceStart;
}
if (askForNewTask) {
checkLocalDirs(fConf.getLocalDirs());
askForNewTask = enoughFreeSpace(localMinSpaceStart);
status.getResourceStatus().setAvailableSpace( getFreeSpace() );
long freeVirtualMem = findFreeVirtualMemory();
LOG.debug("Setting amount of free virtual memory for the new task: " +
freeVirtualMem);
status.getResourceStatus().setFreeVirtualMemory(freeVirtualMem);
status.getResourceStatus().setTotalMemory(maxVirtualMemoryForTasks);
}
//
// Xmit the heartbeat
//
HeartbeatResponse heartbeatResponse = jobClient.heartbeat(status,
justStarted, askForNewTask,
heartbeatResponseId);
//
// The heartbeat got through successfully!
//
heartbeatResponseId = heartbeatResponse.getResponseId();
synchronized (this) {
for (TaskStatus taskStatus : status.getTaskReports()) {
if (taskStatus.getRunState() != TaskStatus.State.RUNNING &&
taskStatus.getRunState() != TaskStatus.State.UNASSIGNED &&
taskStatus.getRunState() != TaskStatus.State.COMMIT_PENDING &&
!taskStatus.inTaskCleanupPhase()) {
if (taskStatus.getIsMap()) {
mapTotal--;
} else {
reduceTotal--;
}
try {
myInstrumentation.completeTask(taskStatus.getTaskID());
} catch (MetricsException me) {
LOG.warn("Caught: " + StringUtils.stringifyException(me));
}
runningTasks.remove(taskStatus.getTaskID());
}
}
// Clear transient status information which should only
// be sent once to the JobTracker
for (TaskInProgress tip: runningTasks.values()) {
tip.getStatus().clearStatus();
}
}
// Force a rebuild of 'status' on the next iteration
status = null;
return heartbeatResponse;
}
/**
* Return the maximum amount of memory available for all tasks on
* this tracker
* @return maximum amount of virtual memory
*/
long getMaxVirtualMemoryForTasks() {
return maxVirtualMemoryForTasks;
}
/**
* Find the minimum amount of virtual memory that would be
* available for a new task.
*
* The minimum amount of virtual memory is computed by looking
* at the maximum amount of virtual memory that is allowed for
* all tasks in the system, as per mapred.tasktracker.tasks.maxmemory,
* and the total amount of maximum virtual memory that can be
* used by all currently running tasks.
*
* @return amount of free virtual memory that can be assured for
* new tasks
*/
private synchronized long findFreeVirtualMemory() {
if (maxVirtualMemoryForTasks == JobConf.DISABLED_VIRTUAL_MEMORY_LIMIT) {
// this will disable picking up tasks based on free memory.
return JobConf.DISABLED_VIRTUAL_MEMORY_LIMIT;
}
long maxMemoryUsed = 0L;
for (TaskInProgress tip: runningTasks.values()) {
// the following task states are one in which the slot is
// still occupied and hence memory of the task should be
// accounted in used memory.
if ((tip.getRunState() == TaskStatus.State.RUNNING)
|| (tip.getRunState() == TaskStatus.State.COMMIT_PENDING)) {
maxMemoryUsed += getMemoryForTask(tip.getJobConf());
}
}
return (maxVirtualMemoryForTasks - maxMemoryUsed);
}
/**
* Return the memory allocated for a TIP.
*
* If the TIP's job has a configured value for the max memory that is
* returned. Else, the default memory that would be assigned for the
* task is returned.
* @param conf
* @return the memory allocated for the TIP.
*/
long getMemoryForTask(JobConf conf) {
long memForTask = conf.getMaxVirtualMemoryForTask();
if (memForTask == JobConf.DISABLED_VIRTUAL_MEMORY_LIMIT) {
memForTask = fConf.getLong("mapred.task.default.maxmemory",
512*1024*1024L);
}
return memForTask;
}
/**
* Check if the jobtracker directed a 'reset' of the tasktracker.
*
* @param actions the directives of the jobtracker for the tasktracker.
* @return <code>true</code> if tasktracker is to be reset,
* <code>false</code> otherwise.
*/
private boolean reinitTaskTracker(TaskTrackerAction[] actions) {
if (actions != null) {
for (TaskTrackerAction action : actions) {
if (action.getActionId() ==
TaskTrackerAction.ActionType.REINIT_TRACKER) {
LOG.info("Recieved RenitTrackerAction from JobTracker");
return true;
}
}
}
return false;
}
/**
* Kill any tasks that have not reported progress in the last X seconds.
*/
private synchronized void markUnresponsiveTasks() throws IOException {
long now = System.currentTimeMillis();
for (TaskInProgress tip: runningTasks.values()) {
if (tip.getRunState() == TaskStatus.State.RUNNING ||
tip.getRunState() == TaskStatus.State.COMMIT_PENDING ||
tip.isCleaningup()) {
// Check the per-job timeout interval for tasks;
// an interval of '0' implies it is never timed-out
long jobTaskTimeout = tip.getTaskTimeout();
if (jobTaskTimeout == 0) {
continue;
}
// Check if the task has not reported progress for a
// time-period greater than the configured time-out
long timeSinceLastReport = now - tip.getLastProgressReport();
if (timeSinceLastReport > jobTaskTimeout && !tip.wasKilled) {
String msg =
"Task " + tip.getTask().getTaskID() + " failed to report status for "
+ (timeSinceLastReport / 1000) + " seconds. Killing!";
LOG.info(tip.getTask().getTaskID() + ": " + msg);
ReflectionUtils.logThreadInfo(LOG, "lost task", 30);
tip.reportDiagnosticInfo(msg);
myInstrumentation.timedoutTask(tip.getTask().getTaskID());
purgeTask(tip, true);
}
}
}
}
/**
* The task tracker is done with this job, so we need to clean up.
* @param action The action with the job
* @throws IOException
*/
private synchronized void purgeJob(KillJobAction action) throws IOException {
JobID jobId = action.getJobID();
LOG.info("Received 'KillJobAction' for job: " + jobId);
RunningJob rjob = null;
synchronized (runningJobs) {
rjob = runningJobs.get(jobId);
}
if (rjob == null) {
LOG.warn("Unknown job " + jobId + " being deleted.");
} else {
synchronized (rjob) {
// Add this tips of this job to queue of tasks to be purged
for (TaskInProgress tip : rjob.tasks) {
tip.jobHasFinished(false);
Task t = tip.getTask();
if (t.isMapTask()) {
indexCache.removeMap(tip.getTask().getTaskID().toString());
}
}
// Delete the job directory for this
// task if the job is done/failed
if (!rjob.keepJobFiles){
directoryCleanupThread.addToQueue(fConf, getLocalFiles(fConf,
getLocalJobDir(rjob.getJobID().toString())));
}
// Remove this job
rjob.tasks.clear();
}
}
synchronized(runningJobs) {
runningJobs.remove(jobId);
}
}
/**
* Remove the tip and update all relevant state.
*
* @param tip {@link TaskInProgress} to be removed.
* @param wasFailure did the task fail or was it killed?
*/
private void purgeTask(TaskInProgress tip, boolean wasFailure)
throws IOException {
if (tip != null) {
LOG.info("About to purge task: " + tip.getTask().getTaskID());
// Remove the task from running jobs,
// removing the job if it's the last task
removeTaskFromJob(tip.getTask().getJobID(), tip);
tip.jobHasFinished(wasFailure);
if (tip.getTask().isMapTask()) {
indexCache.removeMap(tip.getTask().getTaskID().toString());
}
}
}
/** Check if we're dangerously low on disk space
* If so, kill jobs to free up space and make sure
* we don't accept any new tasks
* Try killing the reduce jobs first, since I believe they
* use up most space
* Then pick the one with least progress
*/
private void killOverflowingTasks() throws IOException {
long localMinSpaceKill;
synchronized(this){
localMinSpaceKill = minSpaceKill;
}
if (!enoughFreeSpace(localMinSpaceKill)) {
acceptNewTasks=false;
//we give up! do not accept new tasks until
//all the ones running have finished and they're all cleared up
synchronized (this) {
TaskInProgress killMe = findTaskToKill();
if (killMe!=null) {
String msg = "Tasktracker running out of space." +
" Killing task.";
LOG.info(killMe.getTask().getTaskID() + ": " + msg);
killMe.reportDiagnosticInfo(msg);
purgeTask(killMe, false);
}
}
}
}
/**
* Pick a task to kill to free up space
* @return the task to kill or null, if one wasn't found
*/
private TaskInProgress findTaskToKill() {
TaskInProgress killMe = null;
for (Iterator it = runningTasks.values().iterator(); it.hasNext();) {
TaskInProgress tip = (TaskInProgress) it.next();
if ((tip.getRunState() == TaskStatus.State.RUNNING ||
tip.getRunState() == TaskStatus.State.COMMIT_PENDING) &&
!tip.wasKilled) {
if (killMe == null) {
killMe = tip;
} else if (!tip.getTask().isMapTask()) {
//reduce task, give priority
if (killMe.getTask().isMapTask() ||
(tip.getTask().getProgress().get() <
killMe.getTask().getProgress().get())) {
killMe = tip;
}
} else if (killMe.getTask().isMapTask() &&
tip.getTask().getProgress().get() <
killMe.getTask().getProgress().get()) {
//map task, only add if the progress is lower
killMe = tip;
}
}
}
return killMe;
}
/**
* Check if any of the local directories has enough
* free space (more than minSpace)
*
* If not, do not try to get a new task assigned
* @return
* @throws IOException
*/
private boolean enoughFreeSpace(long minSpace) throws IOException {
if (minSpace == 0) {
return true;
}
return minSpace < getFreeSpace();
}
private long getFreeSpace() throws IOException {
long biggestSeenSoFar = 0;
String[] localDirs = fConf.getLocalDirs();
for (int i = 0; i < localDirs.length; i++) {
DF df = null;
if (localDirsDf.containsKey(localDirs[i])) {
df = localDirsDf.get(localDirs[i]);
} else {
df = new DF(new File(localDirs[i]), fConf);
localDirsDf.put(localDirs[i], df);
}
long availOnThisVol = df.getAvailable();
if (availOnThisVol > biggestSeenSoFar) {
biggestSeenSoFar = availOnThisVol;
}
}
//Should ultimately hold back the space we expect running tasks to use but
//that estimate isn't currently being passed down to the TaskTrackers
return biggestSeenSoFar;
}
/**
* Try to get the size of output for this task.
* Returns -1 if it can't be found.
* @return
*/
long tryToGetOutputSize(TaskAttemptID taskId, JobConf conf) {
try{
TaskInProgress tip;
synchronized(this) {
tip = tasks.get(taskId);
}
if(tip == null)
return -1;
if (!tip.getTask().isMapTask() ||
tip.getRunState() != TaskStatus.State.SUCCEEDED) {
return -1;
}
MapOutputFile mapOutputFile = new MapOutputFile();
mapOutputFile.setJobId(taskId.getJobID());
mapOutputFile.setConf(conf);
Path tmp_output = mapOutputFile.getOutputFile(taskId);
if(tmp_output == null)
return 0;
FileSystem localFS = FileSystem.getLocal(conf);
FileStatus stat = localFS.getFileStatus(tmp_output);
if(stat == null)
return 0;
else
return stat.getLen();
} catch(IOException e) {
LOG.info(e);
return -1;
}
}
private TaskLauncher mapLauncher;
private TaskLauncher reduceLauncher;
public JvmManager getJvmManagerInstance() {
return jvmManager;
}
private void addToTaskQueue(LaunchTaskAction action) {
if (action.getTask().isMapTask()) {
mapLauncher.addToTaskQueue(action);
} else {
reduceLauncher.addToTaskQueue(action);
}
}
private class TaskLauncher extends Thread {
private IntWritable numFreeSlots;
private final int maxSlots;
private List<TaskInProgress> tasksToLaunch;
public TaskLauncher(int numSlots) {
this.maxSlots = numSlots;
this.numFreeSlots = new IntWritable(numSlots);
this.tasksToLaunch = new LinkedList<TaskInProgress>();
setDaemon(true);
setName("TaskLauncher for task");
}
public void addToTaskQueue(LaunchTaskAction action) {
synchronized (tasksToLaunch) {
TaskInProgress tip = registerTask(action, this);
tasksToLaunch.add(tip);
tasksToLaunch.notifyAll();
}
}
public void cleanTaskQueue() {
tasksToLaunch.clear();
}
public void addFreeSlot() {
synchronized (numFreeSlots) {
numFreeSlots.set(numFreeSlots.get() + 1);
assert (numFreeSlots.get() <= maxSlots);
LOG.info("addFreeSlot : current free slots : " + numFreeSlots.get());
numFreeSlots.notifyAll();
}
}
public void run() {
while (!Thread.interrupted()) {
try {
TaskInProgress tip;
synchronized (tasksToLaunch) {
while (tasksToLaunch.isEmpty()) {
tasksToLaunch.wait();
}
//get the TIP
tip = tasksToLaunch.remove(0);
LOG.info("Trying to launch : " + tip.getTask().getTaskID());
}
//wait for a slot to run
synchronized (numFreeSlots) {
while (numFreeSlots.get() == 0) {
numFreeSlots.wait();
}
LOG.info("In TaskLauncher, current free slots : " + numFreeSlots.get()+
" and trying to launch "+tip.getTask().getTaskID());
numFreeSlots.set(numFreeSlots.get() - 1);
assert (numFreeSlots.get() >= 0);
}
synchronized (tip) {
//to make sure that there is no kill task action for this
if (tip.getRunState() != TaskStatus.State.UNASSIGNED &&
tip.getRunState() != TaskStatus.State.FAILED_UNCLEAN &&
tip.getRunState() != TaskStatus.State.KILLED_UNCLEAN) {
//got killed externally while still in the launcher queue
addFreeSlot();
continue;
}
tip.slotTaken = true;
}
//got a free slot. launch the task
startNewTask(tip);
} catch (InterruptedException e) {
return; // ALL DONE
} catch (Throwable th) {
LOG.error("TaskLauncher error " +
StringUtils.stringifyException(th));
}
}
}
}
private TaskInProgress registerTask(LaunchTaskAction action,
TaskLauncher launcher) {
Task t = action.getTask();
LOG.info("LaunchTaskAction (registerTask): " + t.getTaskID() +
" task's state:" + t.getState());
TaskInProgress tip = new TaskInProgress(t, this.fConf, launcher);
synchronized (this) {
tasks.put(t.getTaskID(), tip);
runningTasks.put(t.getTaskID(), tip);
boolean isMap = t.isMapTask();
if (isMap) {
mapTotal++;
} else {
reduceTotal++;
}
}
return tip;
}
/**
* Start a new task.
* All exceptions are handled locally, so that we don't mess up the
* task tracker.
*/
private void startNewTask(TaskInProgress tip) {
try {
localizeJob(tip);
} catch (Throwable e) {
String msg = ("Error initializing " + tip.getTask().getTaskID() +
":\n" + StringUtils.stringifyException(e));
LOG.warn(msg);
tip.reportDiagnosticInfo(msg);
try {
tip.kill(true);
tip.cleanup(true);
} catch (IOException ie2) {
LOG.info("Error cleaning up " + tip.getTask().getTaskID() + ":\n" +
StringUtils.stringifyException(ie2));
}
// Careful!
// This might not be an 'Exception' - don't handle 'Error' here!
if (e instanceof Error) {
throw ((Error) e);
}
}
}
void addToMemoryManager(TaskAttemptID attemptId,
JobConf conf,
String pidFile) {
if (isTaskMemoryManagerEnabled()) {
taskMemoryManager.addTask(attemptId,
getMemoryForTask(conf), pidFile);
}
}
void removeFromMemoryManager(TaskAttemptID attemptId) {
// Remove the entry from taskMemoryManagerThread's data structures.
if (isTaskMemoryManagerEnabled()) {
taskMemoryManager.removeTask(attemptId);
}
}
/**
* The server retry loop.
* This while-loop attempts to connect to the JobTracker. It only
* loops when the old TaskTracker has gone bad (its state is
* stale somehow) and we need to reinitialize everything.
*/
public void run() {
try {
startCleanupThreads();
boolean denied = false;
while (running && !shuttingDown && !denied) {
boolean staleState = false;
try {
// This while-loop attempts reconnects if we get network errors
while (running && !staleState && !shuttingDown && !denied) {
try {
State osState = offerService();
if (osState == State.STALE) {
staleState = true;
} else if (osState == State.DENIED) {
denied = true;
}
} catch (Exception ex) {
if (!shuttingDown) {
LOG.info("Lost connection to JobTracker [" +
jobTrackAddr + "]. Retrying...", ex);
try {
Thread.sleep(5000);
} catch (InterruptedException ie) {
}
}
}
}
} finally {
close();
}
if (shuttingDown) { return; }
LOG.warn("Reinitializing local state");
initialize();
}
if (denied) {
shutdown();
}
} catch (IOException iex) {
LOG.error("Got fatal exception while reinitializing TaskTracker: " +
StringUtils.stringifyException(iex));
return;
}
}
///////////////////////////////////////////////////////
// TaskInProgress maintains all the info for a Task that
// lives at this TaskTracker. It maintains the Task object,
// its TaskStatus, and the TaskRunner.
///////////////////////////////////////////////////////
class TaskInProgress {
Task task;
long lastProgressReport;
StringBuffer diagnosticInfo = new StringBuffer();
private TaskRunner runner;
volatile boolean done = false;
volatile boolean wasKilled = false;
private JobConf defaultJobConf;
private JobConf localJobConf;
private boolean keepFailedTaskFiles;
private boolean alwaysKeepTaskFiles;
private TaskStatus taskStatus;
private long taskTimeout;
private String debugCommand;
private volatile boolean slotTaken = false;
private TaskLauncher launcher;
/**
*/
public TaskInProgress(Task task, JobConf conf) {
this(task, conf, null);
}
public TaskInProgress(Task task, JobConf conf, TaskLauncher launcher) {
this.task = task;
this.launcher = launcher;
this.lastProgressReport = System.currentTimeMillis();
this.defaultJobConf = conf;
localJobConf = null;
taskStatus = TaskStatus.createTaskStatus(task.isMapTask(), task.getTaskID(),
0.0f,
task.getState(),
diagnosticInfo.toString(),
"initializing",
getName(),
task.isTaskCleanupTask() ?
TaskStatus.Phase.CLEANUP :
task.isMapTask()? TaskStatus.Phase.MAP:
TaskStatus.Phase.SHUFFLE,
task.getCounters());
taskTimeout = (10 * 60 * 1000);
}
private void localizeTask(Task task) throws IOException{
Path localTaskDir =
lDirAlloc.getLocalPathForWrite(
TaskTracker.getLocalTaskDir(task.getJobID().toString(),
task.getTaskID().toString(), task.isTaskCleanupTask()),
defaultJobConf );
FileSystem localFs = FileSystem.getLocal(fConf);
if (!localFs.mkdirs(localTaskDir)) {
throw new IOException("Mkdirs failed to create "
+ localTaskDir.toString());
}
// create symlink for ../work if it already doesnt exist
String workDir = lDirAlloc.getLocalPathToRead(
TaskTracker.getLocalJobDir(task.getJobID().toString())
+ Path.SEPARATOR
+ "work", defaultJobConf).toString();
String link = localTaskDir.getParent().toString()
+ Path.SEPARATOR + "work";
File flink = new File(link);
if (!flink.exists())
FileUtil.symLink(workDir, link);
// create the working-directory of the task
Path cwd = lDirAlloc.getLocalPathForWrite(
getLocalTaskDir(task.getJobID().toString(),
task.getTaskID().toString(), task.isTaskCleanupTask())
+ Path.SEPARATOR + MRConstants.WORKDIR,
defaultJobConf);
if (!localFs.mkdirs(cwd)) {
throw new IOException("Mkdirs failed to create "
+ cwd.toString());
}
Path localTaskFile = new Path(localTaskDir, "job.xml");
task.setJobFile(localTaskFile.toString());
localJobConf.set("mapred.local.dir",
fConf.get("mapred.local.dir"));
localJobConf.set("mapred.task.id", task.getTaskID().toString());
keepFailedTaskFiles = localJobConf.getKeepFailedTaskFiles();
task.localizeConfiguration(localJobConf);
List<String[]> staticResolutions = NetUtils.getAllStaticResolutions();
if (staticResolutions != null && staticResolutions.size() > 0) {
StringBuffer str = new StringBuffer();
for (int i = 0; i < staticResolutions.size(); i++) {
String[] hostToResolved = staticResolutions.get(i);
str.append(hostToResolved[0]+"="+hostToResolved[1]);
if (i != staticResolutions.size() - 1) {
str.append(',');
}
}
localJobConf.set("hadoop.net.static.resolutions", str.toString());
}
if (task.isMapTask()) {
debugCommand = localJobConf.getMapDebugScript();
} else {
debugCommand = localJobConf.getReduceDebugScript();
}
String keepPattern = localJobConf.getKeepTaskFilesPattern();
if (keepPattern != null) {
alwaysKeepTaskFiles =
Pattern.matches(keepPattern, task.getTaskID().toString());
} else {
alwaysKeepTaskFiles = false;
}
if (debugCommand != null || localJobConf.getProfileEnabled() ||
alwaysKeepTaskFiles || keepFailedTaskFiles) {
//disable jvm reuse
localJobConf.setNumTasksToExecutePerJvm(1);
}
if (isTaskMemoryManagerEnabled()) {
localJobConf.setBoolean("task.memory.mgmt.enabled", true);
}
OutputStream out = localFs.create(localTaskFile);
try {
localJobConf.writeXml(out);
} finally {
out.close();
}
task.setConf(localJobConf);
}
/**
*/
public Task getTask() {
return task;
}
public TaskRunner getTaskRunner() {
return runner;
}
public synchronized void setJobConf(JobConf lconf){
this.localJobConf = lconf;
keepFailedTaskFiles = localJobConf.getKeepFailedTaskFiles();
taskTimeout = localJobConf.getLong("mapred.task.timeout",
10 * 60 * 1000);
}
public synchronized JobConf getJobConf() {
return localJobConf;
}
/**
*/
public synchronized TaskStatus getStatus() {
taskStatus.setDiagnosticInfo(diagnosticInfo.toString());
if (diagnosticInfo.length() > 0) {
diagnosticInfo = new StringBuffer();
}
return taskStatus;
}
/**
* Kick off the task execution
*/
public synchronized void launchTask() throws IOException {
if (this.taskStatus.getRunState() == TaskStatus.State.UNASSIGNED ||
this.taskStatus.getRunState() == TaskStatus.State.FAILED_UNCLEAN ||
this.taskStatus.getRunState() == TaskStatus.State.KILLED_UNCLEAN) {
localizeTask(task);
if (this.taskStatus.getRunState() == TaskStatus.State.UNASSIGNED) {
this.taskStatus.setRunState(TaskStatus.State.RUNNING);
}
this.runner = task.createRunner(TaskTracker.this, this);
this.runner.start();
this.taskStatus.setStartTime(System.currentTimeMillis());
} else {
LOG.info("Not launching task: " + task.getTaskID() +
" since it's state is " + this.taskStatus.getRunState());
}
}
boolean isCleaningup() {
return this.taskStatus.inTaskCleanupPhase();
}
/**
* The task is reporting its progress
*/
public synchronized void reportProgress(TaskStatus taskStatus)
{
LOG.info(task.getTaskID() + " " + taskStatus.getProgress() +
"% " + taskStatus.getStateString());
// task will report its state as
// COMMIT_PENDING when it is waiting for commit response and
// when it is committing.
// cleanup attempt will report its state as FAILED_UNCLEAN/KILLED_UNCLEAN
if (this.done ||
(this.taskStatus.getRunState() != TaskStatus.State.RUNNING &&
this.taskStatus.getRunState() != TaskStatus.State.COMMIT_PENDING &&
!isCleaningup()) ||
((this.taskStatus.getRunState() == TaskStatus.State.COMMIT_PENDING ||
this.taskStatus.getRunState() == TaskStatus.State.FAILED_UNCLEAN ||
this.taskStatus.getRunState() == TaskStatus.State.KILLED_UNCLEAN) &&
taskStatus.getRunState() == TaskStatus.State.RUNNING)) {
//make sure we ignore progress messages after a task has
//invoked TaskUmbilicalProtocol.done() or if the task has been
//KILLED/FAILED/FAILED_UNCLEAN/KILLED_UNCLEAN
//Also ignore progress update if the state change is from
//COMMIT_PENDING/FAILED_UNCLEAN/KILLED_UNCLEA to RUNNING
LOG.info(task.getTaskID() + " Ignoring status-update since " +
((this.done) ? "task is 'done'" :
("runState: " + this.taskStatus.getRunState()))
);
return;
}
this.taskStatus.statusUpdate(taskStatus);
this.lastProgressReport = System.currentTimeMillis();
}
/**
*/
public long getLastProgressReport() {
return lastProgressReport;
}
/**
*/
public TaskStatus.State getRunState() {
return taskStatus.getRunState();
}
/**
* The task's configured timeout.
*
* @return the task's configured timeout.
*/
public long getTaskTimeout() {
return taskTimeout;
}
/**
* The task has reported some diagnostic info about its status
*/
public synchronized void reportDiagnosticInfo(String info) {
this.diagnosticInfo.append(info);
}
public synchronized void reportNextRecordRange(SortedRanges.Range range) {
this.taskStatus.setNextRecordRange(range);
}
/**
* The task is reporting that it's done running
*/
public synchronized void reportDone() {
if (isCleaningup()) {
if (this.taskStatus.getRunState() == TaskStatus.State.FAILED_UNCLEAN) {
this.taskStatus.setRunState(TaskStatus.State.FAILED);
} else if (this.taskStatus.getRunState() ==
TaskStatus.State.KILLED_UNCLEAN) {
this.taskStatus.setRunState(TaskStatus.State.KILLED);
}
} else {
this.taskStatus.setRunState(TaskStatus.State.SUCCEEDED);
}
this.taskStatus.setProgress(1.0f);
this.taskStatus.setFinishTime(System.currentTimeMillis());
this.done = true;
jvmManager.taskFinished(runner);
runner.signalDone();
LOG.info("Task " + task.getTaskID() + " is done.");
LOG.info("reported output size for " + task.getTaskID() + " was " + taskStatus.getOutputSize());
}
public boolean wasKilled() {
return wasKilled;
}
void reportTaskFinished() {
taskFinished();
releaseSlot();
}
/* State changes:
* RUNNING/COMMIT_PENDING -> FAILED_UNCLEAN/FAILED/KILLED_UNCLEAN/KILLED
* FAILED_UNCLEAN -> FAILED
* KILLED_UNCLEAN -> KILLED
*/
private void setTaskFailState(boolean wasFailure) {
// go FAILED_UNCLEAN -> FAILED and KILLED_UNCLEAN -> KILLED always
if (taskStatus.getRunState() == TaskStatus.State.FAILED_UNCLEAN) {
taskStatus.setRunState(TaskStatus.State.FAILED);
} else if (taskStatus.getRunState() ==
TaskStatus.State.KILLED_UNCLEAN) {
taskStatus.setRunState(TaskStatus.State.KILLED);
} else if (task.isMapOrReduce() &&
taskStatus.getPhase() != TaskStatus.Phase.CLEANUP) {
if (wasFailure) {
taskStatus.setRunState(TaskStatus.State.FAILED_UNCLEAN);
} else {
taskStatus.setRunState(TaskStatus.State.KILLED_UNCLEAN);
}
} else {
if (wasFailure) {
taskStatus.setRunState(TaskStatus.State.FAILED);
} else {
taskStatus.setRunState(TaskStatus.State.KILLED);
}
}
}
/**
* The task has actually finished running.
*/
public void taskFinished() {
long start = System.currentTimeMillis();
//
// Wait until task reports as done. If it hasn't reported in,
// wait for a second and try again.
//
while (!done && (System.currentTimeMillis() - start < WAIT_FOR_DONE)) {
try {
Thread.sleep(1000);
} catch (InterruptedException ie) {
}
}
//
// Change state to success or failure, depending on whether
// task was 'done' before terminating
//
boolean needCleanup = false;
synchronized (this) {
// Remove the task from MemoryManager, if the task SUCCEEDED or FAILED.
// KILLED tasks are removed in method kill(), because Kill
// would result in launching a cleanup attempt before
// TaskRunner returns; if remove happens here, it would remove
// wrong task from memory manager.
if (done || !wasKilled) {
removeFromMemoryManager(task.getTaskID());
}
if (!done) {
if (!wasKilled) {
failures += 1;
setTaskFailState(true);
// call the script here for the failed tasks.
if (debugCommand != null) {
String taskStdout ="";
String taskStderr ="";
String taskSyslog ="";
String jobConf = task.getJobFile();
try {
// get task's stdout file
taskStdout = FileUtil.makeShellPath(
TaskLog.getRealTaskLogFileLocation
(task.getTaskID(), TaskLog.LogName.STDOUT));
// get task's stderr file
taskStderr = FileUtil.makeShellPath(
TaskLog.getRealTaskLogFileLocation
(task.getTaskID(), TaskLog.LogName.STDERR));
// get task's syslog file
taskSyslog = FileUtil.makeShellPath(
TaskLog.getRealTaskLogFileLocation
(task.getTaskID(), TaskLog.LogName.SYSLOG));
} catch(IOException e){
LOG.warn("Exception finding task's stdout/err/syslog files");
}
File workDir = null;
try {
workDir = new File(lDirAlloc.getLocalPathToRead(
TaskTracker.getLocalTaskDir(
task.getJobID().toString(),
task.getTaskID().toString(),
task.isTaskCleanupTask())
+ Path.SEPARATOR + MRConstants.WORKDIR,
localJobConf). toString());
} catch (IOException e) {
LOG.warn("Working Directory of the task " + task.getTaskID() +
"doesnt exist. Caught exception " +
StringUtils.stringifyException(e));
}
// Build the command
File stdout = TaskLog.getRealTaskLogFileLocation(
task.getTaskID(), TaskLog.LogName.DEBUGOUT);
// add pipes program as argument if it exists.
String program ="";
String executable = Submitter.getExecutable(localJobConf);
if ( executable != null) {
try {
program = new URI(executable).getFragment();
} catch (URISyntaxException ur) {
LOG.warn("Problem in the URI fragment for pipes executable");
}
}
String [] debug = debugCommand.split(" ");
Vector<String> vargs = new Vector<String>();
for (String component : debug) {
vargs.add(component);
}
vargs.add(taskStdout);
vargs.add(taskStderr);
vargs.add(taskSyslog);
vargs.add(jobConf);
vargs.add(program);
try {
List<String> wrappedCommand = TaskLog.captureDebugOut
(vargs, stdout);
// run the script.
try {
runScript(wrappedCommand, workDir);
} catch (IOException ioe) {
LOG.warn("runScript failed with: " + StringUtils.
stringifyException(ioe));
}
} catch(IOException e) {
LOG.warn("Error in preparing wrapped debug command");
}
// add all lines of debug out to diagnostics
try {
int num = localJobConf.getInt("mapred.debug.out.lines", -1);
addDiagnostics(FileUtil.makeShellPath(stdout),num,"DEBUG OUT");
} catch(IOException ioe) {
LOG.warn("Exception in add diagnostics!");
}
}
}
taskStatus.setProgress(0.0f);
}
this.taskStatus.setFinishTime(System.currentTimeMillis());
needCleanup = (taskStatus.getRunState() == TaskStatus.State.FAILED ||
taskStatus.getRunState() == TaskStatus.State.FAILED_UNCLEAN ||
taskStatus.getRunState() == TaskStatus.State.KILLED_UNCLEAN ||
taskStatus.getRunState() == TaskStatus.State.KILLED);
}
//
// If the task has failed, or if the task was killAndCleanup()'ed,
// we should clean up right away. We only wait to cleanup
// if the task succeeded, and its results might be useful
// later on to downstream job processing.
//
if (needCleanup) {
removeTaskFromJob(task.getJobID(), this);
}
try {
cleanup(needCleanup);
} catch (IOException ie) {
}
}
/**
* Runs the script given in args
* @param args script name followed by its argumnets
* @param dir current working directory.
* @throws IOException
*/
public void runScript(List<String> args, File dir) throws IOException {
ShellCommandExecutor shexec =
new ShellCommandExecutor(args.toArray(new String[0]), dir);
shexec.execute();
int exitCode = shexec.getExitCode();
if (exitCode != 0) {
throw new IOException("Task debug script exit with nonzero status of "
+ exitCode + ".");
}
}
/**
* Add last 'num' lines of the given file to the diagnostics.
* if num =-1, all the lines of file are added to the diagnostics.
* @param file The file from which to collect diagnostics.
* @param num The number of lines to be sent to diagnostics.
* @param tag The tag is printed before the diagnostics are printed.
*/
public void addDiagnostics(String file, int num, String tag) {
RandomAccessFile rafile = null;
try {
rafile = new RandomAccessFile(file,"r");
int no_lines =0;
String line = null;
StringBuffer tail = new StringBuffer();
tail.append("\n-------------------- "+tag+"---------------------\n");
String[] lines = null;
if (num >0) {
lines = new String[num];
}
while ((line = rafile.readLine()) != null) {
no_lines++;
if (num >0) {
if (no_lines <= num) {
lines[no_lines-1] = line;
}
else { // shift them up
for (int i=0; i<num-1; ++i) {
lines[i] = lines[i+1];
}
lines[num-1] = line;
}
}
else if (num == -1) {
tail.append(line);
tail.append("\n");
}
}
int n = no_lines > num ?num:no_lines;
if (num >0) {
for (int i=0;i<n;i++) {
tail.append(lines[i]);
tail.append("\n");
}
}
if(n!=0)
reportDiagnosticInfo(tail.toString());
} catch (FileNotFoundException fnfe){
LOG.warn("File "+file+ " not found");
} catch (IOException ioe){
LOG.warn("Error reading file "+file);
} finally {
try {
if (rafile != null) {
rafile.close();
}
} catch (IOException ioe) {
LOG.warn("Error closing file "+file);
}
}
}
/**
* We no longer need anything from this task, as the job has
* finished. If the task is still running, kill it and clean up.
*
* @param wasFailure did the task fail, as opposed to was it killed by
* the framework
*/
public void jobHasFinished(boolean wasFailure) throws IOException {
// Kill the task if it is still running
synchronized(this){
if (getRunState() == TaskStatus.State.RUNNING ||
getRunState() == TaskStatus.State.UNASSIGNED ||
getRunState() == TaskStatus.State.COMMIT_PENDING ||
isCleaningup()) {
kill(wasFailure);
}
}
// Cleanup on the finished task
cleanup(true);
}
/**
* Something went wrong and the task must be killed.
* @param wasFailure was it a failure (versus a kill request)?
*/
public synchronized void kill(boolean wasFailure) throws IOException {
if (taskStatus.getRunState() == TaskStatus.State.RUNNING ||
taskStatus.getRunState() == TaskStatus.State.COMMIT_PENDING ||
isCleaningup()) {
wasKilled = true;
if (wasFailure) {
failures += 1;
}
// runner could be null if task-cleanup attempt is not localized yet
if (runner != null) {
runner.kill();
}
setTaskFailState(wasFailure);
} else if (taskStatus.getRunState() == TaskStatus.State.UNASSIGNED) {
if (wasFailure) {
failures += 1;
taskStatus.setRunState(TaskStatus.State.FAILED);
} else {
taskStatus.setRunState(TaskStatus.State.KILLED);
}
}
removeFromMemoryManager(task.getTaskID());
releaseSlot();
}
private synchronized void releaseSlot() {
if (slotTaken) {
if (launcher != null) {
launcher.addFreeSlot();
}
slotTaken = false;
}
}
/**
* The map output has been lost.
*/
private synchronized void mapOutputLost(String failure
) throws IOException {
if (taskStatus.getRunState() == TaskStatus.State.COMMIT_PENDING ||
taskStatus.getRunState() == TaskStatus.State.SUCCEEDED) {
// change status to failure
LOG.info("Reporting output lost:"+task.getTaskID());
taskStatus.setRunState(TaskStatus.State.FAILED);
taskStatus.setProgress(0.0f);
reportDiagnosticInfo("Map output lost, rescheduling: " +
failure);
runningTasks.put(task.getTaskID(), this);
mapTotal++;
} else {
LOG.warn("Output already reported lost:"+task.getTaskID());
}
}
/**
* We no longer need anything from this task. Either the
* controlling job is all done and the files have been copied
* away, or the task failed and we don't need the remains.
* Any calls to cleanup should not lock the tip first.
* cleanup does the right thing- updates tasks in Tasktracker
* by locking tasktracker first and then locks the tip.
*
* if needCleanup is true, the whole task directory is cleaned up.
* otherwise the current working directory of the task
* i.e. &lt;taskid&gt;/work is cleaned up.
*/
void cleanup(boolean needCleanup) throws IOException {
TaskAttemptID taskId = task.getTaskID();
LOG.debug("Cleaning up " + taskId);
synchronized (TaskTracker.this) {
if (needCleanup) {
// see if tasks data structure is holding this tip.
// tasks could hold the tip for cleanup attempt, if cleanup attempt
// got launched before this method.
if (tasks.get(taskId) == this) {
tasks.remove(taskId);
}
}
synchronized (this){
if (alwaysKeepTaskFiles ||
(taskStatus.getRunState() == TaskStatus.State.FAILED &&
keepFailedTaskFiles)) {
return;
}
}
}
synchronized (this) {
try {
// localJobConf could be null if localization has not happened
// then no cleanup will be required.
if (localJobConf == null) {
return;
}
String taskDir = getLocalTaskDir(task.getJobID().toString(),
taskId.toString(), task.isTaskCleanupTask());
if (needCleanup) {
if (runner != null) {
//cleans up the output directory of the task (where map outputs
//and reduce inputs get stored)
runner.close();
}
//We don't delete the workdir
//since some other task (running in the same JVM)
//might be using the dir. The JVM running the tasks would clean
//the workdir per a task in the task process itself.
if (localJobConf.getNumTasksToExecutePerJvm() == 1) {
directoryCleanupThread.addToQueue(defaultJobConf,
getLocalFiles(defaultJobConf,
taskDir));
}
else {
directoryCleanupThread.addToQueue(defaultJobConf,
getLocalFiles(defaultJobConf,
taskDir+"/job.xml"));
}
} else {
if (localJobConf.getNumTasksToExecutePerJvm() == 1) {
directoryCleanupThread.addToQueue(defaultJobConf,
getLocalFiles(defaultJobConf,
taskDir+"/work"));
}
}
} catch (Throwable ie) {
LOG.info("Error cleaning up task runner: " +
StringUtils.stringifyException(ie));
}
}
}
@Override
public boolean equals(Object obj) {
return (obj instanceof TaskInProgress) &&
task.getTaskID().equals
(((TaskInProgress) obj).getTask().getTaskID());
}
@Override
public int hashCode() {
return task.getTaskID().hashCode();
}
}
// ///////////////////////////////////////////////////////////////
// TaskUmbilicalProtocol
/////////////////////////////////////////////////////////////////
/**
* Called upon startup by the child process, to fetch Task data.
*/
public synchronized JvmTask getTask(JVMId jvmId)
throws IOException {
LOG.debug("JVM with ID : " + jvmId + " asked for a task");
if (!jvmManager.isJvmKnown(jvmId)) {
LOG.info("Killing unknown JVM " + jvmId);
return new JvmTask(null, true);
}
RunningJob rjob = runningJobs.get(jvmId.getJobId());
if (rjob == null) { //kill the JVM since the job is dead
LOG.info("Killing JVM " + jvmId + " since job " + jvmId.getJobId() +
" is dead");
jvmManager.killJvm(jvmId);
return new JvmTask(null, true);
}
TaskInProgress tip = jvmManager.getTaskForJvm(jvmId);
if (tip == null) {
return new JvmTask(null, false);
}
if (tasks.get(tip.getTask().getTaskID()) != null) { //is task still present
LOG.info("JVM with ID: " + jvmId + " given task: " +
tip.getTask().getTaskID());
return new JvmTask(tip.getTask(), false);
} else {
LOG.info("Killing JVM with ID: " + jvmId + " since scheduled task: " +
tip.getTask().getTaskID() + " is " + tip.taskStatus.getRunState());
return new JvmTask(null, true);
}
}
/**
* Called periodically to report Task progress, from 0.0 to 1.0.
*/
public synchronized boolean statusUpdate(TaskAttemptID taskid,
TaskStatus taskStatus)
throws IOException {
TaskInProgress tip = tasks.get(taskid);
if (tip != null) {
tip.reportProgress(taskStatus);
return true;
} else {
LOG.warn("Progress from unknown child task: "+taskid);
return false;
}
}
/**
* Called when the task dies before completion, and we want to report back
* diagnostic info
*/
public synchronized void reportDiagnosticInfo(TaskAttemptID taskid, String info) throws IOException {
TaskInProgress tip = tasks.get(taskid);
if (tip != null) {
tip.reportDiagnosticInfo(info);
} else {
LOG.warn("Error from unknown child task: "+taskid+". Ignored.");
}
}
public synchronized void reportNextRecordRange(TaskAttemptID taskid,
SortedRanges.Range range) throws IOException {
TaskInProgress tip = tasks.get(taskid);
if (tip != null) {
tip.reportNextRecordRange(range);
} else {
LOG.warn("reportNextRecordRange from unknown child task: "+taskid+". " +
"Ignored.");
}
}
/** Child checking to see if we're alive. Normally does nothing.*/
public synchronized boolean ping(TaskAttemptID taskid) throws IOException {
return tasks.get(taskid) != null;
}
/**
* Task is reporting that it is in commit_pending
* and it is waiting for the commit Response
*/
public synchronized void commitPending(TaskAttemptID taskid,
TaskStatus taskStatus)
throws IOException {
LOG.info("Task " + taskid + " is in commit-pending," +"" +
" task state:" +taskStatus.getRunState());
statusUpdate(taskid, taskStatus);
reportTaskFinished();
}
/**
* Child checking whether it can commit
*/
public synchronized boolean canCommit(TaskAttemptID taskid) {
return commitResponses.contains(taskid); //don't remove it now
}
/**
* The task is done.
*/
public synchronized void done(TaskAttemptID taskid)
throws IOException {
TaskInProgress tip = tasks.get(taskid);
commitResponses.remove(taskid);
if (tip != null) {
tip.reportDone();
} else {
LOG.warn("Unknown child task done: "+taskid+". Ignored.");
}
}
/**
* A reduce-task failed to shuffle the map-outputs. Kill the task.
*/
public synchronized void shuffleError(TaskAttemptID taskId, String message)
throws IOException {
LOG.fatal("Task: " + taskId + " - Killed due to Shuffle Failure: " + message);
TaskInProgress tip = runningTasks.get(taskId);
tip.reportDiagnosticInfo("Shuffle Error: " + message);
purgeTask(tip, true);
}
/**
* A child task had a local filesystem error. Kill the task.
*/
public synchronized void fsError(TaskAttemptID taskId, String message)
throws IOException {
LOG.fatal("Task: " + taskId + " - Killed due to FSError: " + message);
TaskInProgress tip = runningTasks.get(taskId);
tip.reportDiagnosticInfo("FSError: " + message);
purgeTask(tip, true);
}
public synchronized MapTaskCompletionEventsUpdate getMapCompletionEvents(
JobID jobId, int fromEventId, int maxLocs, TaskAttemptID id)
throws IOException {
TaskCompletionEvent[]mapEvents = TaskCompletionEvent.EMPTY_ARRAY;
synchronized (shouldReset) {
if (shouldReset.remove(id)) {
return new MapTaskCompletionEventsUpdate(mapEvents, true);
}
}
RunningJob rjob;
synchronized (runningJobs) {
rjob = runningJobs.get(jobId);
if (rjob != null) {
synchronized (rjob) {
FetchStatus f = rjob.getFetchStatus();
if (f != null) {
mapEvents = f.getMapEvents(fromEventId, maxLocs);
}
}
}
}
return new MapTaskCompletionEventsUpdate(mapEvents, false);
}
/////////////////////////////////////////////////////
// Called by TaskTracker thread after task process ends
/////////////////////////////////////////////////////
/**
* when you see report task finished, wake up the heartbeat
*/
void reportTaskFinished() {
synchronized(finishedCount) {
finishedCount[0]++;
finishedCount.notify();
}
}
/**
* A completed map task's output has been lost.
*/
public synchronized void mapOutputLost(TaskAttemptID taskid,
String errorMsg) throws IOException {
TaskInProgress tip = tasks.get(taskid);
if (tip != null) {
tip.mapOutputLost(errorMsg);
} else {
LOG.warn("Unknown child with bad map output: "+taskid+". Ignored.");
}
}
/**
* The datastructure for initializing a job
*/
static class RunningJob{
private JobID jobid;
private JobConf jobConf;
// keep this for later use
volatile Set<TaskInProgress> tasks;
boolean localized;
boolean keepJobFiles;
FetchStatus f;
RunningJob(JobID jobid) {
this.jobid = jobid;
localized = false;
tasks = new HashSet<TaskInProgress>();
keepJobFiles = false;
}
JobID getJobID() {
return jobid;
}
void setFetchStatus(FetchStatus f) {
this.f = f;
}
FetchStatus getFetchStatus() {
return f;
}
}
/**
* Get the name for this task tracker.
* @return the string like "tracker_mymachine:50010"
*/
String getName() {
return taskTrackerName;
}
private synchronized List<TaskStatus> cloneAndResetRunningTaskStatuses(
boolean sendCounters) {
List<TaskStatus> result = new ArrayList<TaskStatus>(runningTasks.size());
for(TaskInProgress tip: runningTasks.values()) {
TaskStatus status = tip.getStatus();
status.setIncludeCounters(sendCounters);
status.setOutputSize(tryToGetOutputSize(status.getTaskID(), fConf));
// send counters for finished or failed tasks and commit pending tasks
if (status.getRunState() != TaskStatus.State.RUNNING) {
status.setIncludeCounters(true);
}
result.add((TaskStatus)status.clone());
status.clearStatus();
}
return result;
}
/**
* Get the list of tasks that will be reported back to the
* job tracker in the next heartbeat cycle.
* @return a copy of the list of TaskStatus objects
*/
synchronized List<TaskStatus> getRunningTaskStatuses() {
List<TaskStatus> result = new ArrayList<TaskStatus>(runningTasks.size());
for(TaskInProgress tip: runningTasks.values()) {
result.add(tip.getStatus());
}
return result;
}
/**
* Get the list of stored tasks on this task tracker.
* @return
*/
synchronized List<TaskStatus> getNonRunningTasks() {
List<TaskStatus> result = new ArrayList<TaskStatus>(tasks.size());
for(Map.Entry<TaskAttemptID, TaskInProgress> task: tasks.entrySet()) {
if (!runningTasks.containsKey(task.getKey())) {
result.add(task.getValue().getStatus());
}
}
return result;
}
/**
* Get the list of tasks from running jobs on this task tracker.
* @return a copy of the list of TaskStatus objects
*/
synchronized List<TaskStatus> getTasksFromRunningJobs() {
List<TaskStatus> result = new ArrayList<TaskStatus>(tasks.size());
for (Map.Entry <JobID, RunningJob> item : runningJobs.entrySet()) {
RunningJob rjob = item.getValue();
synchronized (rjob) {
for (TaskInProgress tip : rjob.tasks) {
result.add(tip.getStatus());
}
}
}
return result;
}
/**
* Get the default job conf for this tracker.
*/
JobConf getJobConf() {
return fConf;
}
/**
* Check if the given local directories
* (and parent directories, if necessary) can be created.
* @param localDirs where the new TaskTracker should keep its local files.
* @throws DiskErrorException if all local directories are not writable
*/
private static void checkLocalDirs(String[] localDirs)
throws DiskErrorException {
boolean writable = false;
if (localDirs != null) {
for (int i = 0; i < localDirs.length; i++) {
try {
DiskChecker.checkDir(new File(localDirs[i]));
writable = true;
} catch(DiskErrorException e) {
LOG.warn("Task Tracker local " + e.getMessage());
}
}
}
if (!writable)
throw new DiskErrorException(
"all local directories are not writable");
}
/**
* Is this task tracker idle?
* @return has this task tracker finished and cleaned up all of its tasks?
*/
public synchronized boolean isIdle() {
return tasks.isEmpty() && tasksToCleanup.isEmpty();
}
/**
* Start the TaskTracker, point toward the indicated JobTracker
*/
public static void main(String argv[]) throws Exception {
StringUtils.startupShutdownMessage(TaskTracker.class, argv, LOG);
if (argv.length != 0) {
System.out.println("usage: TaskTracker");
System.exit(-1);
}
try {
JobConf conf=new JobConf();
// enable the server to track time spent waiting on locks
ReflectionUtils.setContentionTracing
(conf.getBoolean("tasktracker.contention.tracking", false));
new TaskTracker(conf).run();
} catch (Throwable e) {
LOG.error("Can not start task tracker because "+
StringUtils.stringifyException(e));
System.exit(-1);
}
}
/**
* This class is used in TaskTracker's Jetty to serve the map outputs
* to other nodes.
*/
public static class MapOutputServlet extends HttpServlet {
private static final int MAX_BYTES_TO_READ = 64 * 1024;
@Override
public void doGet(HttpServletRequest request,
HttpServletResponse response
) throws ServletException, IOException {
String mapId = request.getParameter("map");
String reduceId = request.getParameter("reduce");
String jobId = request.getParameter("job");
if (jobId == null) {
throw new IOException("job parameter is required");
}
if (mapId == null || reduceId == null) {
throw new IOException("map and reduce parameters are required");
}
ServletContext context = getServletContext();
int reduce = Integer.parseInt(reduceId);
byte[] buffer = new byte[MAX_BYTES_TO_READ];
// true iff IOException was caused by attempt to access input
boolean isInputException = true;
OutputStream outStream = null;
FSDataInputStream mapOutputIn = null;
IFileInputStream checksumInputStream = null;
long totalRead = 0;
ShuffleServerMetrics shuffleMetrics = (ShuffleServerMetrics)
context.getAttribute("shuffleServerMetrics");
TaskTracker tracker =
(TaskTracker) context.getAttribute("task.tracker");
try {
shuffleMetrics.serverHandlerBusy();
outStream = response.getOutputStream();
JobConf conf = (JobConf) context.getAttribute("conf");
LocalDirAllocator lDirAlloc =
(LocalDirAllocator)context.getAttribute("localDirAllocator");
FileSystem fileSys =
(FileSystem) context.getAttribute("local.file.system");
// Index file
Path indexFileName = lDirAlloc.getLocalPathToRead(
TaskTracker.getIntermediateOutputDir(jobId, mapId)
+ "/file.out.index", conf);
// Map-output file
Path mapOutputFileName = lDirAlloc.getLocalPathToRead(
TaskTracker.getIntermediateOutputDir(jobId, mapId)
+ "/file.out", conf);
/**
* Read the index file to get the information about where
* the map-output for the given reducer is available.
*/
IndexRecord info =
tracker.indexCache.getIndexInformation(mapId, reduce,indexFileName);
final long startOffset = info.startOffset;
final long rawPartLength = info.rawLength;
final long partLength = info.partLength;
//set the custom "Raw-Map-Output-Length" http header to
//the raw (decompressed) length
response.setHeader(RAW_MAP_OUTPUT_LENGTH, Long.toString(rawPartLength));
//set the custom "Map-Output-Length" http header to
//the actual number of bytes being transferred
response.setHeader(MAP_OUTPUT_LENGTH,
Long.toString(partLength));
//use the same buffersize as used for reading the data from disk
response.setBufferSize(MAX_BYTES_TO_READ);
/**
* Read the data from the sigle map-output file and
* send it to the reducer.
*/
//open the map-output file
FileSystem rfs = ((LocalFileSystem)fileSys).getRaw();
mapOutputIn = rfs.open(mapOutputFileName);
// TODO: Remove this after a 'fix' for HADOOP-3647
// The clever trick here to reduce the impact of the extra seek for
// logging the first key/value lengths is to read the lengths before
// the second seek for the actual shuffle. The second seek is almost
// a no-op since it is very short (go back length of two VInts) and the
// data is almost guaranteed to be in the filesystem's buffers.
// WARN: This won't work for compressed map-outputs!
int firstKeyLength = 0;
int firstValueLength = 0;
if (partLength > 0) {
mapOutputIn.seek(startOffset);
firstKeyLength = WritableUtils.readVInt(mapOutputIn);
firstValueLength = WritableUtils.readVInt(mapOutputIn);
}
//seek to the correct offset for the reduce
mapOutputIn.seek(startOffset);
checksumInputStream = new IFileInputStream(mapOutputIn,partLength);
int len = checksumInputStream.readWithChecksum(buffer, 0,
partLength < MAX_BYTES_TO_READ
? (int)partLength : MAX_BYTES_TO_READ);
while (len > 0) {
try {
shuffleMetrics.outputBytes(len);
outStream.write(buffer, 0, len);
outStream.flush();
} catch (IOException ie) {
isInputException = false;
throw ie;
}
totalRead += len;
if (totalRead == partLength) break;
len = checksumInputStream.readWithChecksum(buffer, 0,
(partLength - totalRead) < MAX_BYTES_TO_READ
? (int)(partLength - totalRead) : MAX_BYTES_TO_READ);
}
LOG.info("Sent out " + totalRead + " bytes for reduce: " + reduce +
" from map: " + mapId + " given " + partLength + "/" +
rawPartLength + " from " + startOffset + " with (" +
firstKeyLength + ", " + firstValueLength + ")");
} catch (IOException ie) {
Log log = (Log) context.getAttribute("log");
String errorMsg = ("getMapOutput(" + mapId + "," + reduceId +
") failed :\n"+
StringUtils.stringifyException(ie));
log.warn(errorMsg);
if (isInputException) {
tracker.mapOutputLost(TaskAttemptID.forName(mapId), errorMsg);
}
response.sendError(HttpServletResponse.SC_GONE, errorMsg);
shuffleMetrics.failedOutput();
throw ie;
} finally {
if (checksumInputStream != null) {
checksumInputStream.close();
}
shuffleMetrics.serverHandlerFree();
if (ClientTraceLog.isInfoEnabled()) {
ClientTraceLog.info(String.format(MR_CLIENTTRACE_FORMAT,
request.getLocalAddr() + ":" + request.getLocalPort(),
request.getRemoteAddr() + ":" + request.getRemotePort(),
totalRead, "MAPRED_SHUFFLE", mapId));
}
}
outStream.close();
shuffleMetrics.successOutput();
}
}
// get the full paths of the directory in all the local disks.
private Path[] getLocalFiles(JobConf conf, String subdir) throws IOException{
String[] localDirs = conf.getLocalDirs();
Path[] paths = new Path[localDirs.length];
FileSystem localFs = FileSystem.getLocal(conf);
for (int i = 0; i < localDirs.length; i++) {
paths[i] = new Path(localDirs[i], subdir);
paths[i] = paths[i].makeQualified(localFs);
}
return paths;
}
int getMaxCurrentMapTasks() {
return maxCurrentMapTasks;
}
int getMaxCurrentReduceTasks() {
return maxCurrentReduceTasks;
}
/**
* Is the TaskMemoryManager Enabled on this system?
* @return true if enabled, false otherwise.
*/
public boolean isTaskMemoryManagerEnabled() {
return taskMemoryManagerEnabled;
}
public TaskMemoryManagerThread getTaskMemoryManager() {
return taskMemoryManager;
}
private void setTaskMemoryManagerEnabledFlag() {
if (!ProcfsBasedProcessTree.isAvailable()) {
LOG.info("ProcessTree implementation is missing on this system. "
+ "TaskMemoryManager is disabled.");
taskMemoryManagerEnabled = false;
return;
}
Long tasksMaxMem = getMaxVirtualMemoryForTasks();
if (tasksMaxMem == JobConf.DISABLED_VIRTUAL_MEMORY_LIMIT) {
LOG.info("TaskTracker's tasksMaxMem is not set. TaskMemoryManager is "
+ "disabled.");
taskMemoryManagerEnabled = false;
return;
}
taskMemoryManagerEnabled = true;
}
/**
* Clean-up the task that TaskMemoryMangerThread requests to do so.
* @param tid
* @param diagnosticMsg
*/
synchronized void cleanUpOverMemoryTask(TaskAttemptID tid,
String diagnosticMsg) {
TaskInProgress tip = runningTasks.get(tid);
if (tip != null) {
tip.reportDiagnosticInfo(diagnosticMsg);
try {
purgeTask(tip, true); // Marking it as failure.
} catch (IOException ioe) {
LOG.warn("Couldn't purge the task of " + tid + ". Error : " + ioe);
}
}
}
}