List of usage examples for org.apache.hadoop.io DataOutputBuffer DataOutputBuffer
public DataOutputBuffer()
From source file:org.commoncrawl.util.TextBytes.java
License:Open Source License
public static void main(String[] args) { // run some tests on the new code String aTestString = new String("A Test Strnig"); // convert it to bytes byte bytes[] = aTestString.getBytes(); // over allocate an array byte overAllocated[] = new byte[bytes.length * 2]; // copy source System.arraycopy(bytes, 0, overAllocated, bytes.length, bytes.length); // now allocate a TextBytes TextBytes textBytes = new TextBytes(); // set the overallocated buffer as the backing store textBytes.set(overAllocated, bytes.length, bytes.length); // convert it to string first String toString = textBytes.toString(); // validate equal to original Assert.assertTrue(aTestString.equals(toString)); // ok now write it to output buffer DataOutputBuffer outputBuffer = new DataOutputBuffer(); // write string try {/* ww w.j av a2s.co m*/ textBytes.write(outputBuffer); // read length DataInputBuffer inputBuffer = new DataInputBuffer(); inputBuffer.reset(outputBuffer.getData(), 0, outputBuffer.size()); int encodedLength = WritableUtils.readVInt(inputBuffer); // validate arrays match ... Assert.assertTrue(encodedLength == bytes.length); Assert.assertEquals(WritableComparator.compareBytes(bytes, 0, bytes.length, outputBuffer.getData(), inputBuffer.getPosition(), outputBuffer.getLength() - inputBuffer.getPosition()), 0); // ok reset input buffer again ... inputBuffer.reset(outputBuffer.getData(), 0, outputBuffer.size()); // read in fields textBytes.readFields(inputBuffer); // ok see if we are not using the original backing store ... Assert.assertTrue(textBytes.getBytes() != overAllocated); // validate buffers match to original Assert.assertEquals(WritableComparator.compareBytes(bytes, 0, bytes.length, textBytes.getBytes(), textBytes.getOffset(), textBytes.getLength()), 0); } catch (IOException e) { e.printStackTrace(); } }
From source file:org.commoncrawl.util.TimeSeriesDataFile.java
License:Open Source License
/** append a record to the file ... * //from w ww . j a v a 2 s . com * @param key * @param value * @throws IOException */ public synchronized long appendRecordToLogFile(long key, Writable value) throws IOException { LogFileHeader header = new LogFileHeader(); boolean preExistingHeader = fileName.exists(); RandomAccessFile file = new RandomAccessFile(fileName, "rw"); long recordPositionOut = -1; try { if (preExistingHeader) { long headerOffset = readLogFileHeader(file, header); if (header._writePos == 0) { recordPositionOut = headerOffset; } else { recordPositionOut = header._writePos; } // seelk to appropriate write position file.seek(recordPositionOut); } else { recordPositionOut = writeLogFileHeader(file, header); } DataOutputBuffer buffer = new DataOutputBuffer(); // write out sync bytes ... buffer.writeInt(SyncBytes); // write out placeholder for record length buffer.writeInt(0); // write out placeholder for crc buffer.writeLong(0); // write out key + value to buffer WritableUtils.writeVLong(buffer, key); // write out value ... value.write(buffer); // write out trailing record size (4 bytes sync + 4 bytes record length + 4 bytes crc + key/value buffer + buffer.writeInt(buffer.getLength()); // reset crc crc.reset(); //calc crc crc.update(buffer.getData(), RECORD_HEADER_LENGTH, buffer.getLength() - RECORD_HEADER_LENGTH); // ok fix up record ... // write out record length // total length - sync bytes(4) - record length(4), at offset 4 writeInt(buffer.getLength() - 8, 4, buffer.getData()); // and write out crc // at offset 8 (after sync(4) and length(4) writeLong(crc.getValue(), 8, buffer.getData()); // and then the data file.write(buffer.getData(), 0, buffer.getLength()); // now update header ... header._itemCount += 1; header._writePos = file.getFilePointer(); header._lastRecordLength = buffer.getLength() - 4; header._lastRecordKey = key; // now write out header anew ... writeLogFileHeader(file, header); } finally { if (file != null) { file.close(); } } return recordPositionOut; }
From source file:org.commoncrawl.util.Tuples.java
License:Open Source License
static void validateTextTuple() { // validate tuple code IntAndTwoTextByteTuples tuple1 = new IntAndTwoTextByteTuples(); IntAndTwoTextByteTuples tuple2 = new IntAndTwoTextByteTuples(); tuple1.setIntValue(1);/*from w w w . j a va 2 s . c om*/ tuple2.setIntValue(1); tuple1.setTextValueBytes(new TextBytes("AAAAA")); tuple2.setTextValueBytes(new TextBytes("AAAAA")); tuple1.setSecondTextValueBytes(new TextBytes("AAAAA")); tuple2.setSecondTextValueBytes(new TextBytes("AAAAB")); // compare the two Assert.assertTrue(tuple1.compareTo(tuple2) == -1); tuple1.setTextValueBytes(new TextBytes("BAAAA")); Assert.assertTrue(tuple1.compareTo(tuple2) == 1); tuple2.setIntValue(2); Assert.assertTrue(tuple1.compareTo(tuple2) == -1); // ok restore ... tuple1.setTextValueBytes(new TextBytes("AAAAA")); tuple2.setTextValueBytes(new TextBytes("AAAAA")); tuple1.setSecondTextValueBytes(new TextBytes("AAAAA")); tuple2.setSecondTextValueBytes(new TextBytes("AAAAB")); DataOutputBuffer outputBuffer = new DataOutputBuffer(); try { tuple1.write(outputBuffer); tuple2.write(outputBuffer); IntAndTwoTextByteTuples tuple3 = new IntAndTwoTextByteTuples(); IntAndTwoTextByteTuples tuple4 = new IntAndTwoTextByteTuples(); DataInputBuffer inputBuffer = new DataInputBuffer(); inputBuffer.reset(outputBuffer.getData(), 0, outputBuffer.getLength()); tuple3.readFields(inputBuffer); tuple4.readFields(inputBuffer); Assert.assertTrue(tuple3.compareTo(tuple1) == 0); Assert.assertTrue(tuple4.compareTo(tuple2) == 0); } catch (IOException e) { // TODO Auto-generated catch block e.printStackTrace(); } }
From source file:org.commoncrawl.util.Tuples.java
License:Open Source License
static void validateBufferTuple() { // run some tests on the new code String aTestString = new String("A Test Strnig"); // convert it to bytes byte bytes[] = aTestString.getBytes(); // over allocate an array byte overAllocated[] = new byte[bytes.length * 2]; // copy source System.arraycopy(bytes, 0, overAllocated, bytes.length, bytes.length); IntBufferTuple tuple1 = new IntBufferTuple(); IntBufferTuple tuple2 = new IntBufferTuple(); tuple1.setIntValue(1);/*from www.j a v a2 s. c o m*/ tuple2.setIntValue(1); tuple1.getBuffer().set(overAllocated, bytes.length, bytes.length); tuple2.getBuffer().set(overAllocated, bytes.length, bytes.length); Assert.assertTrue(tuple1.compareTo(tuple2) == 0); DataOutputBuffer outputBuffer = new DataOutputBuffer(); try { tuple1.write(outputBuffer); tuple2.write(outputBuffer); DataInputBuffer inputBuffer = new DataInputBuffer(); inputBuffer.reset(outputBuffer.getData(), 0, outputBuffer.getLength()); tuple1.readFields(inputBuffer); tuple2.readFields(inputBuffer); Assert.assertTrue(tuple1.compareTo(tuple2) == 0); DataOutputBuffer outputBuffer2 = new DataOutputBuffer(); tuple1.write(outputBuffer2); tuple2.write(outputBuffer2); Assert.assertTrue(WritableComparator.compareBytes(outputBuffer.getData(), 0, outputBuffer.getLength(), outputBuffer2.getData(), 0, outputBuffer2.getLength()) == 0); } catch (IOException e) { e.printStackTrace(); } }
From source file:org.conan.myhadoop02.mr.yarntest.Client.java
License:Apache License
/** * Main run function for the client// w w w . j av a 2 s. c om * @return true if application completed successfully * @throws IOException * @throws YarnException */ public boolean run() throws IOException, YarnException { LOG.info("Running Client"); yarnClient.start(); YarnClusterMetrics clusterMetrics = yarnClient.getYarnClusterMetrics(); LOG.info("Got Cluster metric info from ASM" + ", numNodeManagers=" + clusterMetrics.getNumNodeManagers()); List<NodeReport> clusterNodeReports = yarnClient.getNodeReports(NodeState.RUNNING); LOG.info("Got Cluster node info from ASM"); for (NodeReport node : clusterNodeReports) { LOG.info("Got node report from ASM for" + ", nodeId=" + node.getNodeId() + ", nodeAddress" + node.getHttpAddress() + ", nodeRackName" + node.getRackName() + ", nodeNumContainers" + node.getNumContainers()); } QueueInfo queueInfo = yarnClient.getQueueInfo(this.amQueue); LOG.info("Queue info" + ", queueName=" + queueInfo.getQueueName() + ", queueCurrentCapacity=" + queueInfo.getCurrentCapacity() + ", queueMaxCapacity=" + queueInfo.getMaximumCapacity() + ", queueApplicationCount=" + queueInfo.getApplications().size() + ", queueChildQueueCount=" + queueInfo.getChildQueues().size()); List<QueueUserACLInfo> listAclInfo = yarnClient.getQueueAclsInfo(); for (QueueUserACLInfo aclInfo : listAclInfo) { for (QueueACL userAcl : aclInfo.getUserAcls()) { LOG.info("User ACL Info for Queue" + ", queueName=" + aclInfo.getQueueName() + ", userAcl=" + userAcl.name()); } } if (domainId != null && domainId.length() > 0 && toCreateDomain) { prepareTimelineDomain(); } // Get a new application id YarnClientApplication app = yarnClient.createApplication(); GetNewApplicationResponse appResponse = app.getNewApplicationResponse(); // TODO get min/max resource capabilities from RM and change memory ask if needed // If we do not have min/max, we may not be able to correctly request // the required resources from the RM for the app master // Memory ask has to be a multiple of min and less than max. // Dump out information about cluster capability as seen by the resource manager int maxMem = appResponse.getMaximumResourceCapability().getMemory(); LOG.info("Max mem capabililty of resources in this cluster " + maxMem); // A resource ask cannot exceed the max. if (amMemory > maxMem) { LOG.info("AM memory specified above max threshold of cluster. Using max value." + ", specified=" + amMemory + ", max=" + maxMem); amMemory = maxMem; } int maxVCores = appResponse.getMaximumResourceCapability().getVirtualCores(); LOG.info("Max virtual cores capabililty of resources in this cluster " + maxVCores); if (amVCores > maxVCores) { LOG.info("AM virtual cores specified above max threshold of cluster. " + "Using max value." + ", specified=" + amVCores + ", max=" + maxVCores); amVCores = maxVCores; } // set the application name ApplicationSubmissionContext appContext = app.getApplicationSubmissionContext(); ApplicationId appId = appContext.getApplicationId(); appContext.setKeepContainersAcrossApplicationAttempts(keepContainers); appContext.setApplicationName(appName); if (attemptFailuresValidityInterval >= 0) { appContext.setAttemptFailuresValidityInterval(attemptFailuresValidityInterval); } // set local resources for the application master // local files or archives as needed // In this scenario, the jar file for the application master is part of the local resources Map<String, LocalResource> localResources = new HashMap<String, LocalResource>(); LOG.info("Copy App Master jar from local filesystem and add to local environment"); // Copy the application master jar to the filesystem // Create a local resource to point to the destination jar path FileSystem fs = FileSystem.get(conf); addToLocalResources(fs, appMasterJar, appMasterJarPath, appId.toString(), localResources, null); // Set the log4j properties if needed if (!log4jPropFile.isEmpty()) { addToLocalResources(fs, log4jPropFile, log4jPath, appId.toString(), localResources, null); } // The shell script has to be made available on the final container(s) // where it will be executed. // To do this, we need to first copy into the filesystem that is visible // to the yarn framework. // We do not need to set this as a local resource for the application // master as the application master does not need it. String hdfsShellScriptLocation = ""; long hdfsShellScriptLen = 0; long hdfsShellScriptTimestamp = 0; if (!shellScriptPath.isEmpty()) { Path shellSrc = new Path(shellScriptPath); String shellPathSuffix = appName + "/" + appId.toString() + "/" + SCRIPT_PATH; Path shellDst = new Path(fs.getHomeDirectory(), shellPathSuffix); fs.copyFromLocalFile(false, true, shellSrc, shellDst); hdfsShellScriptLocation = shellDst.toUri().toString(); FileStatus shellFileStatus = fs.getFileStatus(shellDst); hdfsShellScriptLen = shellFileStatus.getLen(); hdfsShellScriptTimestamp = shellFileStatus.getModificationTime(); } if (!shellCommand.isEmpty()) { addToLocalResources(fs, null, shellCommandPath, appId.toString(), localResources, shellCommand); } if (shellArgs.length > 0) { addToLocalResources(fs, null, shellArgsPath, appId.toString(), localResources, StringUtils.join(shellArgs, " ")); } // Set the necessary security tokens as needed //amContainer.setContainerTokens(containerToken); // Set the env variables to be setup in the env where the application master will be run LOG.info("Set the environment for the application master"); Map<String, String> env = new HashMap<String, String>(); // put location of shell script into env // using the env info, the application master will create the correct local resource for the // eventual containers that will be launched to execute the shell scripts env.put(DSConstants.DISTRIBUTEDSHELLSCRIPTLOCATION, hdfsShellScriptLocation); env.put(DSConstants.DISTRIBUTEDSHELLSCRIPTTIMESTAMP, Long.toString(hdfsShellScriptTimestamp)); env.put(DSConstants.DISTRIBUTEDSHELLSCRIPTLEN, Long.toString(hdfsShellScriptLen)); if (domainId != null && domainId.length() > 0) { env.put(DSConstants.DISTRIBUTEDSHELLTIMELINEDOMAIN, domainId); } // Add AppMaster.jar location to classpath // At some point we should not be required to add // the hadoop specific classpaths to the env. // It should be provided out of the box. // For now setting all required classpaths including // the classpath to "." for the application jar StringBuilder classPathEnv = new StringBuilder(Environment.CLASSPATH.$$()) .append(ApplicationConstants.CLASS_PATH_SEPARATOR).append("./*"); for (String c : conf.getStrings(YarnConfiguration.YARN_APPLICATION_CLASSPATH, YarnConfiguration.DEFAULT_YARN_CROSS_PLATFORM_APPLICATION_CLASSPATH)) { classPathEnv.append(ApplicationConstants.CLASS_PATH_SEPARATOR); classPathEnv.append(c.trim()); } classPathEnv.append(ApplicationConstants.CLASS_PATH_SEPARATOR).append("./log4j.properties"); // add the runtime classpath needed for tests to work if (conf.getBoolean(YarnConfiguration.IS_MINI_YARN_CLUSTER, false)) { classPathEnv.append(':'); classPathEnv.append(System.getProperty("java.class.path")); } env.put("CLASSPATH", classPathEnv.toString()); // Set the necessary command to execute the application master Vector<CharSequence> vargs = new Vector<CharSequence>(30); // Set java executable command LOG.info("Setting up app master command"); vargs.add(Environment.JAVA_HOME.$$() + "/bin/java"); // Set Xmx based on am memory size vargs.add("-Xmx" + amMemory + "m"); // Set class name vargs.add(appMasterMainClass); // Set params for Application Master vargs.add("--container_memory " + String.valueOf(containerMemory)); vargs.add("--container_vcores " + String.valueOf(containerVirtualCores)); vargs.add("--num_containers " + String.valueOf(numContainers)); if (null != nodeLabelExpression) { appContext.setNodeLabelExpression(nodeLabelExpression); } vargs.add("--priority " + String.valueOf(shellCmdPriority)); for (Map.Entry<String, String> entry : shellEnv.entrySet()) { vargs.add("--shell_env " + entry.getKey() + "=" + entry.getValue()); } if (debugFlag) { vargs.add("--debug"); } vargs.add("1>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/AppMaster.stdout"); vargs.add("2>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/AppMaster.stderr"); // Get final commmand StringBuilder command = new StringBuilder(); for (CharSequence str : vargs) { command.append(str).append(" "); } LOG.info("Completed setting up app master command " + command.toString()); List<String> commands = new ArrayList<String>(); commands.add(command.toString()); // Set up the container launch context for the application master ContainerLaunchContext amContainer = ContainerLaunchContext.newInstance(localResources, env, commands, null, null, null); // Set up resource type requirements // For now, both memory and vcores are supported, so we set memory and // vcores requirements Resource capability = Resource.newInstance(amMemory, amVCores); appContext.setResource(capability); // Service data is a binary blob that can be passed to the application // Not needed in this scenario // amContainer.setServiceData(serviceData); // Setup security tokens if (UserGroupInformation.isSecurityEnabled()) { // Note: Credentials class is marked as LimitedPrivate for HDFS and MapReduce Credentials credentials = new Credentials(); String tokenRenewer = conf.get(YarnConfiguration.RM_PRINCIPAL); if (tokenRenewer == null || tokenRenewer.length() == 0) { throw new IOException("Can't get Master Kerberos principal for the RM to use as renewer"); } // For now, only getting tokens for the default file-system. final Token<?> tokens[] = fs.addDelegationTokens(tokenRenewer, credentials); if (tokens != null) { for (Token<?> token : tokens) { LOG.info("Got dt for " + fs.getUri() + "; " + token); } } DataOutputBuffer dob = new DataOutputBuffer(); credentials.writeTokenStorageToStream(dob); ByteBuffer fsTokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength()); amContainer.setTokens(fsTokens); } appContext.setAMContainerSpec(amContainer); // Set the priority for the application master // TODO - what is the range for priority? how to decide? Priority pri = Priority.newInstance(amPriority); appContext.setPriority(pri); // Set the queue to which this application is to be submitted in the RM appContext.setQueue(amQueue); // Submit the application to the applications manager // SubmitApplicationResponse submitResp = applicationsManager.submitApplication(appRequest); // Ignore the response as either a valid response object is returned on success // or an exception thrown to denote some form of a failure LOG.info("Submitting application to ASM"); yarnClient.submitApplication(appContext); // TODO // Try submitting the same request again // app submission failure? // Monitor the application return monitorApplication(appId); }
From source file:org.dknight.app.ApplicationMaster.java
License:Apache License
/** * Main run function for the application master * * @throws YarnException//from w w w .j a v a 2 s .c o m * @throws IOException */ @SuppressWarnings({ "unchecked" }) public boolean run() throws YarnException, IOException { LOG.info("Starting ApplicationMaster"); Credentials credentials = UserGroupInformation.getCurrentUser().getCredentials(); DataOutputBuffer dob = new DataOutputBuffer(); credentials.writeTokenStorageToStream(dob); // Now remove the AM->RM token so that containers cannot access it. Iterator<Token<?>> iter = credentials.getAllTokens().iterator(); while (iter.hasNext()) { Token<?> token = iter.next(); if (token.getKind().equals(AMRMTokenIdentifier.KIND_NAME)) { iter.remove(); } } allTokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength()); AMRMClientAsync.CallbackHandler allocListener = new RMCallbackHandler(); amRMClient = AMRMClientAsync.createAMRMClientAsync(1000, allocListener); amRMClient.init(conf); amRMClient.start(); containerListener = createNMCallbackHandler(); nmClientAsync = new NMClientAsyncImpl(containerListener); nmClientAsync.init(conf); nmClientAsync.start(); yarnClient = YarnClient.createYarnClient(); yarnClient.init(conf); yarnClient.start(); // Setup local RPC Server to accept status requests directly from clients // TODO need to setup a protocol for client to be able to communicate to // the RPC server // TODO use the rpc port info to register with the RM for the client to // send requests to this app master // Register self with ResourceManager // This will start heartbeating to the RM appMasterHostname = NetUtils.getHostname(); RegisterApplicationMasterResponse response = amRMClient.registerApplicationMaster(appMasterHostname, appMasterRpcPort, appMasterTrackingUrl); // Dump out information about cluster capability as seen by the // resource manager int maxMem = response.getMaximumResourceCapability().getMemory(); LOG.info("Max mem capabililty of resources in this cluster " + maxMem); // A resource ask cannot exceed the max. if (containerMemory > maxMem) { LOG.info("Container memory specified above max threshold of cluster." + " Using max value." + ", specified=" + containerMemory + ", max=" + maxMem); containerMemory = maxMem; } // Setup ask for containers from RM // Send request for containers to RM // Until we get our fully allocated quota, we keep on polling RM for // containers // Keep looping until all the containers are launched and shell script // executed on them ( regardless of success/failure). List<NodeReport> nodeReports = null; if (yarnClient != null) { try { nodeReports = yarnClient.getNodeReports(NodeState.RUNNING); for (NodeReport node : nodeReports) { LOG.info("Got node report from ASM for" + ", nodeId=" + node.getNodeId() + ", nodeAddress" + node.getHttpAddress() + ", nodeRackName" + node.getRackName() + ", nodeNumContainers" + node.getNumContainers()); } } catch (YarnException e) { e.printStackTrace(); //To change body of catch statement use File | Settings | File Templates. } catch (IOException e) { e.printStackTrace(); //To change body of catch statement use File | Settings | File Templates. } } for (int i = 0; i < numTotalContainers; ++i) { ContainerRequest containerAsk = null; if (CollectionUtils.isNotEmpty(nodeReports)) { NodeReport requestNode = nodeReports.get(Math.min(i, nodeReports.size() - 1)); String[] requsetHosts = new String[] { requestNode.getNodeId().getHost() }; LOG.info("Ask container at the host list:" + requsetHosts); containerAsk = setupContainerAskForRM(requsetHosts); } else { LOG.info("Ask container with ANY host"); containerAsk = setupContainerAskForRM(null); } amRMClient.addContainerRequest(containerAsk); } numRequestedContainers.set(numTotalContainers); while (!done && (numCompletedContainers.get() != numTotalContainers)) { try { Thread.sleep(200); } catch (InterruptedException ex) { } } finish(); return success; }
From source file:org.dknight.app.Client.java
License:Apache License
/** * Main run function for the client/*from www . jav a 2 s . c o m*/ * @return true if application completed successfully * @throws IOException * @throws YarnException */ public boolean run() throws IOException, YarnException { LOG.info("Running Client"); yarnClient.start(); YarnClusterMetrics clusterMetrics = yarnClient.getYarnClusterMetrics(); LOG.info("Got Cluster metric info from ASM" + ", numNodeManagers=" + clusterMetrics.getNumNodeManagers()); List<NodeReport> clusterNodeReports = yarnClient.getNodeReports(NodeState.RUNNING); LOG.info("Got Cluster node info from ASM"); for (NodeReport node : clusterNodeReports) { LOG.info("Got node report from ASM for" + ", nodeId=" + node.getNodeId() + ", nodeAddress" + node.getHttpAddress() + ", nodeRackName" + node.getRackName() + ", nodeNumContainers" + node.getNumContainers()); } QueueInfo queueInfo = yarnClient.getQueueInfo(this.amQueue); LOG.info("Queue info" + ", queueName=" + queueInfo.getQueueName() + ", queueCurrentCapacity=" + queueInfo.getCurrentCapacity() + ", queueMaxCapacity=" + queueInfo.getMaximumCapacity() + ", queueApplicationCount=" + queueInfo.getApplications().size() + ", queueChildQueueCount=" + queueInfo.getChildQueues().size()); List<QueueUserACLInfo> listAclInfo = yarnClient.getQueueAclsInfo(); for (QueueUserACLInfo aclInfo : listAclInfo) { for (QueueACL userAcl : aclInfo.getUserAcls()) { LOG.info("User ACL Info for Queue" + ", queueName=" + aclInfo.getQueueName() + ", userAcl=" + userAcl.name()); } } // Get a new application id YarnClientApplication app = yarnClient.createApplication(); GetNewApplicationResponse appResponse = app.getNewApplicationResponse(); // TODO get min/max resource capabilities from RM and change memory ask if needed // If we do not have min/max, we may not be able to correctly request // the required resources from the RM for the app master // Memory ask has to be a multiple of min and less than max. // Dump out information about cluster capability as seen by the resource manager int maxMem = appResponse.getMaximumResourceCapability().getMemory(); LOG.info("Max mem capabililty of resources in this cluster " + maxMem); // A resource ask cannot exceed the max. if (amMemory > maxMem) { LOG.info("AM memory specified above max threshold of cluster. Using max value." + ", specified=" + amMemory + ", max=" + maxMem); amMemory = maxMem; } // set the application name ApplicationSubmissionContext appContext = app.getApplicationSubmissionContext(); ApplicationId appId = appContext.getApplicationId(); appContext.setApplicationName(appName); // Set up the container launch context for the application master ContainerLaunchContext amContainer = Records.newRecord(ContainerLaunchContext.class); // set local resources for the application master // local files or archives as needed // In this scenario, the jar file for the application master is part of the local resources Map<String, LocalResource> localResources = new HashMap<String, LocalResource>(); LOG.info("Copy App Master jar from local filesystem and add to local environment"); // Copy the application master jar to the filesystem // Create a local resource to point to the destination jar path FileSystem fs = FileSystem.get(conf); Path src = new Path(appMasterJar); String pathSuffix = appName + "/" + appId.getId() + "/AppMaster.jar"; Path dst = new Path(fs.getHomeDirectory(), pathSuffix); fs.copyFromLocalFile(false, true, src, dst); FileStatus destStatus = fs.getFileStatus(dst); LocalResource amJarRsrc = Records.newRecord(LocalResource.class); // Set the type of resource - file or archive // archives are untarred at destination // we don't need the jar file to be untarred for now amJarRsrc.setType(LocalResourceType.FILE); // Set visibility of the resource // Setting to most private option amJarRsrc.setVisibility(LocalResourceVisibility.APPLICATION); // Set the resource to be copied over amJarRsrc.setResource(ConverterUtils.getYarnUrlFromPath(dst)); // Set timestamp and length of file so that the framework // can do basic sanity checks for the local resource // after it has been copied over to ensure it is the same // resource the client intended to use with the application amJarRsrc.setTimestamp(destStatus.getModificationTime()); amJarRsrc.setSize(destStatus.getLen()); localResources.put("AppMaster.jar", amJarRsrc); String confXMLFSPath = ""; { File clusterConfXML = new File("cluster-conf.xml"); conf.writeXml(new FileOutputStream(clusterConfXML)); Path confSrc = new Path(clusterConfXML.getAbsolutePath()); String confPathSuffix = appName + "/" + appId.getId() + "/cluster-conf.xml"; Path confDst = new Path(fs.getHomeDirectory(), confPathSuffix); fs.copyFromLocalFile(false, true, confSrc, confDst); FileStatus confFileStatus = fs.getFileStatus(confDst); LocalResource confRsrc = Records.newRecord(LocalResource.class); confRsrc.setType(LocalResourceType.FILE); confRsrc.setVisibility(LocalResourceVisibility.APPLICATION); confRsrc.setResource(ConverterUtils.getYarnUrlFromURI(confDst.toUri())); confRsrc.setSize(confFileStatus.getLen()); confRsrc.setTimestamp(confFileStatus.getModificationTime()); localResources.put("cluster-conf.xml", confRsrc); confXMLFSPath = confDst.toUri().getPath(); } // Set the log4j properties if needed if (!log4jPropFile.isEmpty()) { Path log4jSrc = new Path(log4jPropFile); Path log4jDst = new Path(fs.getHomeDirectory(), "log4j.props"); fs.copyFromLocalFile(false, true, log4jSrc, log4jDst); FileStatus log4jFileStatus = fs.getFileStatus(log4jDst); LocalResource log4jRsrc = Records.newRecord(LocalResource.class); log4jRsrc.setType(LocalResourceType.FILE); log4jRsrc.setVisibility(LocalResourceVisibility.APPLICATION); log4jRsrc.setResource(ConverterUtils.getYarnUrlFromURI(log4jDst.toUri())); log4jRsrc.setTimestamp(log4jFileStatus.getModificationTime()); log4jRsrc.setSize(log4jFileStatus.getLen()); localResources.put("log4j.properties", log4jRsrc); } // The shell script has to be made available on the final container(s) // where it will be executed. // To do this, we need to first copy into the filesystem that is visible // to the yarn framework. // We do not need to set this as a local resource for the application // master as the application master does not need it. String hdfsShellScriptLocation = ""; long hdfsShellScriptLen = 0; long hdfsShellScriptTimestamp = 0; if (!shellScriptPath.isEmpty()) { Path shellSrc = new Path(shellScriptPath); String shellPathSuffix = appName + "/" + appId.getId() + "/ExecShellScript.sh"; Path shellDst = new Path(fs.getHomeDirectory(), shellPathSuffix); fs.copyFromLocalFile(false, true, shellSrc, shellDst); hdfsShellScriptLocation = shellDst.toUri().toString(); FileStatus shellFileStatus = fs.getFileStatus(shellDst); hdfsShellScriptLen = shellFileStatus.getLen(); hdfsShellScriptTimestamp = shellFileStatus.getModificationTime(); } // Set local resource info into app master container launch context amContainer.setLocalResources(localResources); // Set the necessary security tokens as needed //amContainer.setContainerTokens(containerToken); // Set the env variables to be setup in the env where the application master will be run LOG.info("Set the environment for the application master"); Map<String, String> env = new HashMap<String, String>(); // put location of shell script into env // using the env info, the application master will create the correct local resource for the // eventual containers that will be launched to execute the shell scripts env.put(DSConstants.DISTRIBUTEDSHELLSCRIPTLOCATION, hdfsShellScriptLocation); env.put(DSConstants.DISTRIBUTEDSHELLSCRIPTTIMESTAMP, Long.toString(hdfsShellScriptTimestamp)); env.put(DSConstants.DISTRIBUTEDSHELLSCRIPTLEN, Long.toString(hdfsShellScriptLen)); env.put(DSConstants.CLUSTER_CONF_XML_PATH, confXMLFSPath); // Add AppMaster.jar location to classpath // At some point we should not be required to add // the hadoop specific classpaths to the env. // It should be provided out of the box. // For now setting all required classpaths including // the classpath to "." for the application jar StringBuilder classPathEnv = new StringBuilder(Environment.CLASSPATH.$()).append(File.pathSeparatorChar) .append("./*"); for (String c : conf.getStrings(YarnConfiguration.YARN_APPLICATION_CLASSPATH, YarnConfiguration.DEFAULT_YARN_APPLICATION_CLASSPATH)) { classPathEnv.append(File.pathSeparatorChar); classPathEnv.append(c.trim()); } classPathEnv.append(File.pathSeparatorChar).append("./log4j.properties"); // add the runtime classpath needed for tests to work if (conf.getBoolean(YarnConfiguration.IS_MINI_YARN_CLUSTER, false)) { classPathEnv.append(':'); classPathEnv.append(System.getProperty("java.class.path")); } env.put("CLASSPATH", classPathEnv.toString()); amContainer.setEnvironment(env); // Set the necessary command to execute the application master Vector<CharSequence> vargs = new Vector<CharSequence>(30); // Set java executable command LOG.info("Setting up app master command"); vargs.add(Environment.JAVA_HOME.$() + "/bin/java"); // Set Xmx based on am memory size vargs.add("-Xmx" + amMemory + "m"); // if (debugFlag) { // vargs.add("-Xdebug -Xrunjdwp:transport=dt_socket,address=9998,server=y,suspend=y"); // } // Set class name vargs.add(appMasterMainClass); // Set params for Application Master vargs.add("--container_memory " + String.valueOf(containerMemory)); vargs.add("--num_containers " + String.valueOf(numContainers)); vargs.add("--priority " + String.valueOf(shellCmdPriority)); if (!shellCommand.isEmpty()) { vargs.add("--shell_command " + shellCommand + ""); } if (!shellArgs.isEmpty()) { vargs.add("--shell_args " + shellArgs + ""); } for (Map.Entry<String, String> entry : shellEnv.entrySet()) { vargs.add("--shell_env " + entry.getKey() + "=" + entry.getValue()); } if (debugFlag) { vargs.add("--debug"); } vargs.add("1>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/AppMaster.stdout"); vargs.add("2>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/AppMaster.stderr"); // Get final commmand StringBuilder command = new StringBuilder(); for (CharSequence str : vargs) { command.append(str).append(" "); } LOG.info("Completed setting up app master command " + command.toString()); List<String> commands = new ArrayList<String>(); commands.add(command.toString()); amContainer.setCommands(commands); // Set up resource type requirements // For now, only memory is supported so we set memory requirements Resource capability = Records.newRecord(Resource.class); capability.setMemory(amMemory); appContext.setResource(capability); // Service data is a binary blob that can be passed to the application // Not needed in this scenario // amContainer.setServiceData(serviceData); // Setup security tokens if (UserGroupInformation.isSecurityEnabled()) { Credentials credentials = new Credentials(); String tokenRenewer = conf.get(YarnConfiguration.RM_PRINCIPAL); if (tokenRenewer == null || tokenRenewer.length() == 0) { throw new IOException("Can't get Master Kerberos principal for the RM to use as renewer"); } // For now, only getting tokens for the default file-system. final Token<?> tokens[] = fs.addDelegationTokens(tokenRenewer, credentials); if (tokens != null) { for (Token<?> token : tokens) { LOG.info("Got dt for " + fs.getUri() + "; " + token); } } DataOutputBuffer dob = new DataOutputBuffer(); credentials.writeTokenStorageToStream(dob); ByteBuffer fsTokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength()); amContainer.setTokens(fsTokens); } appContext.setAMContainerSpec(amContainer); // Set the priority for the application master Priority pri = Records.newRecord(Priority.class); // TODO - what is the range for priority? how to decide? pri.setPriority(amPriority); appContext.setPriority(pri); // Set the queue to which this application is to be submitted in the RM appContext.setQueue(amQueue); // Submit the application to the applications manager // SubmitApplicationResponse submitResp = applicationsManager.submitApplication(appRequest); // Ignore the response as either a valid response object is returned on success // or an exception thrown to denote some form of a failure LOG.info("Submitting application to ASM"); yarnClient.submitApplication(appContext); // TODO // Try submitting the same request again // app submission failure? // Monitor the application return monitorApplication(appId); }
From source file:org.goldenorb.io.InputSplitAllocator.java
License:Apache License
/** * This method gets the raw splits and calls another method to assign them. * //from w ww .j a v a 2 s. c o m * @returns Map */ @SuppressWarnings({ "deprecation", "rawtypes", "unchecked" }) public Map<OrbPartitionMember, List<RawSplit>> assignInputSplits() { List<RawSplit> rawSplits = null; JobConf job = new JobConf(orbConf); LOG.debug(orbConf.getJobNumber().toString()); JobContext jobContext = new JobContext(job, new JobID(orbConf.getJobNumber(), 0)); org.apache.hadoop.mapreduce.InputFormat<?, ?> input; try { input = ReflectionUtils.newInstance(jobContext.getInputFormatClass(), orbConf); List<org.apache.hadoop.mapreduce.InputSplit> splits = input.getSplits(jobContext); rawSplits = new ArrayList<RawSplit>(splits.size()); DataOutputBuffer buffer = new DataOutputBuffer(); SerializationFactory factory = new SerializationFactory(orbConf); Serializer serializer = factory.getSerializer(splits.get(0).getClass()); serializer.open(buffer); for (int i = 0; i < splits.size(); i++) { buffer.reset(); serializer.serialize(splits.get(i)); RawSplit rawSplit = new RawSplit(); rawSplit.setClassName(splits.get(i).getClass().getName()); rawSplit.setDataLength(splits.get(i).getLength()); rawSplit.setBytes(buffer.getData(), 0, buffer.getLength()); rawSplit.setLocations(splits.get(i).getLocations()); rawSplits.add(rawSplit); } } catch (ClassNotFoundException e) { e.printStackTrace(); throw new RuntimeException(e); } catch (IOException e) { e.printStackTrace(); throw new RuntimeException(e); } catch (InterruptedException e) { e.printStackTrace(); throw new RuntimeException(e); } return assignInputSplits(rawSplits); }
From source file:org.hdl.caffe.yarn.app.ApplicationMaster.java
License:Apache License
/** * Main run function for the application master * * @throws YarnException/*ww w . j a v a 2s .c om*/ * @throws IOException */ @SuppressWarnings({ "unchecked" }) public void run() throws YarnException, IOException, InterruptedException { LOG.info("Starting ApplicationMaster"); // Note: Credentials, Token, UserGroupInformation, DataOutputBuffer class // are marked as LimitedPrivate Credentials credentials = UserGroupInformation.getCurrentUser().getCredentials(); DataOutputBuffer dob = new DataOutputBuffer(); credentials.writeTokenStorageToStream(dob); // Now remove the AM->RM token so that containers cannot access it. Iterator<Token<?>> iter = credentials.getAllTokens().iterator(); LOG.info("Executing with tokens:"); while (iter.hasNext()) { Token<?> token = iter.next(); LOG.info(token); if (token.getKind().equals(AMRMTokenIdentifier.KIND_NAME)) { iter.remove(); } } allTokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength()); // Create appSubmitterUgi and add original tokens to it String appSubmitterUserName = System.getenv(ApplicationConstants.Environment.USER.name()); appSubmitterUgi = UserGroupInformation.createRemoteUser(appSubmitterUserName); appSubmitterUgi.addCredentials(credentials); AMRMClientAsync.AbstractCallbackHandler allocListener = new RMCallbackHandler(); amRMClient = AMRMClientAsync.createAMRMClientAsync(1000, allocListener); amRMClient.init(conf); amRMClient.start(); containerListener = createNMCallbackHandler(); nmClientAsync = new NMClientAsyncImpl(containerListener); nmClientAsync.init(conf); nmClientAsync.start(); appMasterHostname = System.getenv(Environment.NM_HOST.name()); CaffeApplicationRpcServer rpcServer = new CaffeApplicationRpcServer(appMasterHostname, new RpcForClient()); appMasterRpcPort = rpcServer.getRpcPort(); rpcServer.startRpcServiceThread(); // Register self with ResourceManager // This will start heartbeating to the RM RegisterApplicationMasterResponse response = amRMClient.registerApplicationMaster(appMasterHostname, appMasterRpcPort, appMasterTrackingUrl); // Dump out information about cluster capability as seen by the // resource manager long maxMem = response.getMaximumResourceCapability().getMemorySize(); LOG.info("Max mem capability of resources in this cluster " + maxMem); int maxVCores = response.getMaximumResourceCapability().getVirtualCores(); LOG.info("Max vcores capability of resources in this cluster " + maxVCores); // A resource ask cannot exceed the max. if (containerMemory > maxMem) { LOG.info("Container memory specified above max threshold of cluster." + " Using max value." + ", specified=" + containerMemory + ", max=" + maxMem); containerMemory = maxMem; } if (containerVirtualCores > maxVCores) { LOG.info("Container virtual cores specified above max threshold of cluster." + " Using max value." + ", specified=" + containerVirtualCores + ", max=" + maxVCores); containerVirtualCores = maxVCores; } List<Container> previousAMRunningContainers = response.getContainersFromPreviousAttempts(); LOG.info(appAttemptID + " received " + previousAMRunningContainers.size() + " previous attempts' running containers on AM registration."); for (Container container : previousAMRunningContainers) { launchedContainers.add(container.getId()); } numAllocatedContainers.addAndGet(previousAMRunningContainers.size()); int numTotalContainersToRequest = numTotalContainers - previousAMRunningContainers.size(); // Setup ask for containers from RM // Send request for containers to RM // Until we get our fully allocated quota, we keep on polling RM for // containers // Keep looping until all the containers are launched and shell script // executed on them ( regardless of success/failure). for (int i = 0; i < numTotalContainersToRequest; ++i) { ContainerRequest containerAsk = setupContainerAskForRM(); amRMClient.addContainerRequest(containerAsk); } numRequestedContainers.set(numTotalContainers); }
From source file:org.hdl.caffe.yarn.app.Client.java
License:Apache License
/** * Main run function for the client/*w w w .j av a 2s . c o m*/ * * @return true if application completed successfully * @throws IOException * @throws YarnException */ public boolean run() throws IOException, YarnException { yarnClient.start(); YarnClusterMetrics clusterMetrics = yarnClient.getYarnClusterMetrics(); LOG.info("Got Cluster metric info from ASM" + ", numNodeManagers=" + clusterMetrics.getNumNodeManagers()); List<NodeReport> clusterNodeReports = yarnClient.getNodeReports(NodeState.RUNNING); LOG.info("Got Cluster node info from ASM"); for (NodeReport node : clusterNodeReports) { LOG.info("Got node report from ASM for" + ", nodeId=" + node.getNodeId() + ", nodeAddress=" + node.getHttpAddress() + ", nodeRackName=" + node.getRackName() + ", nodeNumContainers=" + node.getNumContainers()); } QueueInfo queueInfo = yarnClient.getQueueInfo(this.amQueue); LOG.info("Queue info" + ", queueName=" + queueInfo.getQueueName() + ", queueCurrentCapacity=" + queueInfo.getCurrentCapacity() + ", queueMaxCapacity=" + queueInfo.getMaximumCapacity() + ", queueApplicationCount=" + queueInfo.getApplications().size() + ", queueChildQueueCount=" + queueInfo.getChildQueues().size()); List<QueueUserACLInfo> listAclInfo = yarnClient.getQueueAclsInfo(); for (QueueUserACLInfo aclInfo : listAclInfo) { for (QueueACL userAcl : aclInfo.getUserAcls()) { LOG.info("User ACL Info for Queue" + ", queueName=" + aclInfo.getQueueName() + ", userAcl=" + userAcl.name()); } } // Get a new application id YarnClientApplication app = yarnClient.createApplication(); GetNewApplicationResponse appResponse = app.getNewApplicationResponse(); long maxMem = appResponse.getMaximumResourceCapability().getMemorySize(); LOG.info("Max mem capability of resources in this cluster " + maxMem); if (amMemory > maxMem) { LOG.info("AM memory specified above max threshold of cluster. Using max value." + ", specified=" + amMemory + ", max=" + maxMem); amMemory = maxMem; } int maxVCores = appResponse.getMaximumResourceCapability().getVirtualCores(); LOG.info("Max virtual cores capability of resources in this cluster " + maxVCores); if (amVCores > maxVCores) { LOG.info("AM virtual cores specified above max threshold of cluster. " + "Using max value." + ", specified=" + amVCores + ", max=" + maxVCores); amVCores = maxVCores; } ApplicationSubmissionContext appContext = app.getApplicationSubmissionContext(); ApplicationId appId = appContext.getApplicationId(); appContext.setApplicationName(appName); if (attemptFailuresValidityInterval >= 0) { appContext.setAttemptFailuresValidityInterval(attemptFailuresValidityInterval); } Set<String> tags = new HashSet<String>(); appContext.setApplicationTags(tags); Map<String, LocalResource> localResources = new HashMap<String, LocalResource>(); CaffeAmContainer CaffeAmContainer = new CaffeAmContainer(this); // Copy the application jar to the filesystem FileSystem fs = FileSystem.get(conf); String dstJarPath = copyLocalFileToDfs(fs, appId.toString(), appMasterJar, CaffeContainer.SERVER_JAR_PATH); CaffeAmContainer.addToLocalResources(fs, new Path(dstJarPath), CaffeAmContainer.APPMASTER_JAR_PATH, localResources); Map<String, String> env = CaffeAmContainer.setJavaEnv(conf); env.put("LD_LIBRARY_PATH", "/root/CaffeOnSpark/caffe-public/distribute/lib:/root/CaffeOnSpark/caffe-distri/distribute/lib"); if (null != nodeLabelExpression) { appContext.setNodeLabelExpression(nodeLabelExpression); } StringBuilder command = CaffeAmContainer.makeCommands(amMemory, appMasterMainClass, containerMemory, containerVirtualCores, processorNum, dstJarPath, containerRetryOptions, train, solver, feature, label, model, output, connection); LOG.info("AppMaster command: " + command.toString()); List<String> commands = new ArrayList<String>(); commands.add(command.toString()); ContainerLaunchContext amContainer = ContainerLaunchContext.newInstance(localResources, env, commands, null, null, null); Resource capability = Resource.newInstance(amMemory, amVCores); appContext.setResource(capability); // Service data is a binary blob that can be passed to the application // Not needed in this scenario // amContainer.setServiceData(serviceData); // Setup security tokens if (UserGroupInformation.isSecurityEnabled()) { // Note: Credentials class is marked as LimitedPrivate for HDFS and MapReduce Credentials credentials = new Credentials(); String tokenRenewer = YarnClientUtils.getRmPrincipal(conf); if (tokenRenewer == null || tokenRenewer.length() == 0) { throw new IOException("Can't get Master Kerberos principal for the RM to use as renewer"); } // For now, only getting tokens for the default file-system. final Token<?> tokens[] = fs.addDelegationTokens(tokenRenewer, credentials); if (tokens != null) { for (Token<?> token : tokens) { LOG.info("Got dt for " + fs.getUri() + "; " + token); } } DataOutputBuffer dob = new DataOutputBuffer(); credentials.writeTokenStorageToStream(dob); ByteBuffer fsTokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength()); amContainer.setTokens(fsTokens); } appContext.setAMContainerSpec(amContainer); // Set the priority for the application master Priority pri = Priority.newInstance(amPriority); appContext.setPriority(pri); appContext.setQueue(amQueue); LOG.info("Submitting application to ASM"); yarnClient.submitApplication(appContext); handleSignal(appId); return monitorApplication(appId); }