List of usage examples for org.apache.hadoop.yarn.client.util YarnClientUtils getRmPrincipal
public static String getRmPrincipal(Configuration conf) throws IOException
From source file:com.github.hdl.tensorflow.yarn.app.Client.java
License:Apache License
/** * Main run function for the client// www. j a va2 s. 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(); // TODO get min/max resource capabilities from RM and change memory ask if needed 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>(); TFAmContainer tfAmContainer = new TFAmContainer(this); // Copy the application jar to the filesystem FileSystem fs = FileSystem.get(conf); String dstJarPath = copyLocalFileToDfs(fs, appId.toString(), appMasterJar, TFContainer.SERVER_JAR_PATH); tfAmContainer.addToLocalResources(fs, new Path(dstJarPath), TFAmContainer.APPMASTER_JAR_PATH, localResources); String jniSoDfsPath = ""; if (jniSoFile != null && !jniSoFile.equals("")) { jniSoDfsPath = copyLocalFileToDfs(fs, appId.toString(), jniSoFile, "libbridge.so"); } // Set the log4j properties if needed /* if (!log4jPropFile.isEmpty()) { tfAmContainer.addToLocalResources(fs, log4jPropFile, log4jPath, appId.toString(), localResources, null); }*/ // Set the necessary security tokens as needed //amContainer.setContainerTokens(containerToken); Map<String, String> env = tfAmContainer.setJavaEnv(conf); if (null != nodeLabelExpression) { appContext.setNodeLabelExpression(nodeLabelExpression); } StringBuilder command = tfAmContainer.makeCommands(amMemory, appMasterMainClass, containerMemory, containerVirtualCores, workerNum, psNum, dstJarPath, containerRetryOptions, jniSoDfsPath); 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 // TODO - what is the range for priority? how to decide? 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); }
From source file:org.hdl.caffe.yarn.app.Client.java
License:Apache License
/** * Main run function for the client/* w ww .j a v a 2 s .co 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); }