List of usage examples for org.apache.hadoop.fs Path toUri
public URI toUri()
From source file:com.bigjob.Client.java
License:Apache License
/** * Main run function for the client/*from w w w . j a va 2s . 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 (RM)" + ", 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; } 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.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 // if (dfsUrl!=null && dfsUrl.equals("")==false){ // conf.set("fs.defaultFS", dfsUrl); // } FileSystem fs = FileSystem.get(conf); addToLocalResources(fs, appMasterJar, appMasterJarPath, appId.getId(), localResources, null); // Set the log4j properties if needed if (!log4jPropFile.isEmpty()) { addToLocalResources(fs, log4jPropFile, log4jPath, appId.getId(), 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.getId() + "/" + (Shell.WINDOWS ? windowBatPath : linuxShellPath); 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.getId(), localResources, shellCommand); } if (shellArgs.length > 0) { addToLocalResources(fs, null, shellArgsPath, appId.getId(), localResources, StringUtils.join(shellArgs, " ")); } // 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)); // 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"); // 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)); 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()); amContainer.setCommands(commands); // Set up resource type requirements // For now, both memory and vcores are supported, so we set memory and // vcores requirements Resource capability = Records.newRecord(Resource.class); capability.setMemory(amMemory); capability.setVirtualCores(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()) { 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); System.out.println("ApplicationId:" + appId); return true; }
From source file:com.bigjob.Client.java
License:Apache License
private void addToLocalResources(FileSystem fs, String fileSrcPath, String fileDstPath, int appId, Map<String, LocalResource> localResources, String resources) throws IOException { String suffix = appName + "/" + appId + "/" + fileDstPath; Path dst = new Path(fs.getHomeDirectory(), suffix); LOG.debug("HDFS Destination for Script: " + dst.toString()); if (fileSrcPath == null) { FSDataOutputStream ostream = null; try {/*w w w .j a v a 2 s. c om*/ ostream = FileSystem.create(fs, dst, new FsPermission((short) 0710)); ostream.writeUTF(resources); } finally { IOUtils.closeQuietly(ostream); } } else { fs.copyFromLocalFile(new Path(fileSrcPath), dst); } FileStatus scFileStatus = fs.getFileStatus(dst); LocalResource scRsrc = LocalResource.newInstance(ConverterUtils.getYarnUrlFromURI(dst.toUri()), LocalResourceType.FILE, LocalResourceVisibility.APPLICATION, scFileStatus.getLen(), scFileStatus.getModificationTime()); localResources.put(fileDstPath, scRsrc); }
From source file:com.bigstep.datalake.DLFileSystem.java
License:Apache License
/** * This computes a fully qualified path from a relative path. * Makes sure all paths start with a /data_lake/dl prefix. * If none is provided it will prepend the home directory. * @param path//w w w . j a v a2s .c o m * @return the resulting path string. */ @Override public Path makeQualified(Path path) { URI pathURI = path.toUri(); String datalakelRelativePath = pathURI.getPath().startsWith("/data_lake/dl") ? pathURI.getPath() : this.getHomeDirectory() + pathURI.getPath(); return super.makeQualified(new Path(datalakelRelativePath)); }
From source file:com.blackberry.logdriver.admin.LogMaintenance.java
License:Apache License
@Override public int run(String[] args) throws Exception { Configuration conf = getConf(); // If run by Oozie, then load the Oozie conf too if (System.getProperty("oozie.action.conf.xml") != null) { conf.addResource(new URL("file://" + System.getProperty("oozie.action.conf.xml"))); }//from w ww. j a va 2 s. c o m // For some reason, Oozie needs some options to be set in system instead of // in the confiuration. So copy the configs over. { Iterator<Entry<String, String>> i = conf.iterator(); while (i.hasNext()) { Entry<String, String> next = i.next(); System.setProperty(next.getKey(), next.getValue()); } } if (args.length < 3) { printUsage(); return 1; } String userName = args[0]; String dcNumber = args[1]; String service = args[2]; String date = null; String hour = null; if (args.length >= 4) { date = args[3]; } if (args.length >= 5) { hour = args[4]; } // Set from environment variables String mergeJobPropertiesFile = getConfOrEnv(conf, "MERGEJOB_CONF"); String filterJobPropertiesFile = getConfOrEnv(conf, "FILTERJOB_CONF"); String daysBeforeArchive = getConfOrEnv(conf, "DAYS_BEFORE_ARCHIVE"); String daysBeforeDelete = getConfOrEnv(conf, "DAYS_BEFORE_DELETE"); String maxConcurrentMR = getConfOrEnv(conf, "MAX_CONCURRENT_MR", "-1"); String zkConnectString = getConfOrEnv(conf, "ZK_CONNECT_STRING"); String logdir = getConfOrEnv(conf, "logdriver.logdir.name"); boolean resetOrphanedJobs = Boolean.parseBoolean(getConfOrEnv(conf, "reset.orphaned.jobs", "true")); String rootDir = getConfOrEnv(conf, "service.root.dir"); String maxTotalMR = getConfOrEnv(conf, "MAX_TOTAL_MR", "-1"); boolean doMerge = true; boolean doArchive = true; boolean doDelete = true; if (zkConnectString == null) { LOG.error("ZK_CONNECT_STRING is not set. Exiting."); return 1; } if (mergeJobPropertiesFile == null) { LOG.info("MERGEJOB_CONF is not set. Not merging."); doMerge = false; } if (filterJobPropertiesFile == null) { LOG.info("FILTERJOB_CONF is not set. Not archiving."); doArchive = false; } if (daysBeforeArchive == null) { LOG.info("DAYS_BEFORE_ARCHIVE is not set. Not archiving."); doArchive = false; } if (doArchive && Integer.parseInt(daysBeforeArchive) < 0) { LOG.info("DAYS_BEFORE_ARCHIVE is negative. Not archiving."); doArchive = false; } if (daysBeforeDelete == null) { LOG.info("DAYS_BEFORE_DELETE is not set. Not deleting."); doDelete = false; } if (doDelete && Integer.parseInt(daysBeforeDelete) < 0) { LOG.info("DAYS_BEFORE_DELETE is negative. Not deleting."); doDelete = false; } if (logdir == null) { LOG.info("LOGDRIVER_LOGDIR_NAME is not set. Using default value of 'logs'."); logdir = "logs"; } if (rootDir == null) { LOG.info("SERVICE_ROOT_DIR is not set. Using default value of 'service'."); rootDir = "/service"; } // We can hang if this fails. So make sure we abort if it fails. fs = null; try { fs = FileSystem.get(conf); fs.exists(new Path("/")); // Test if it works. } catch (IOException e) { LOG.error("Error getting filesystem.", e); return 1; } // Create the LockUtil instance lockUtil = new LockUtil(zkConnectString); // Now it's safe to create our Job Runner JobRunner jobRunner = new JobRunner(Integer.parseInt(maxConcurrentMR), Integer.parseInt(maxTotalMR)); Thread jobRunnerThread = new Thread(jobRunner); jobRunnerThread.setName("JobRunner"); jobRunnerThread.setDaemon(false); jobRunnerThread.start(); // Figure out what date we start filters on. String filterCutoffDate = ""; if (doArchive) { Calendar cal = Calendar.getInstance(); cal.add(Calendar.DAY_OF_MONTH, Integer.parseInt("-" + daysBeforeArchive)); filterCutoffDate = String.format("%04d%02d%02d%02d", cal.get(Calendar.YEAR), (cal.get(Calendar.MONTH) + 1), cal.get(Calendar.DAY_OF_MONTH), cal.get(Calendar.HOUR_OF_DAY)); LOG.info("Archiving logs from before {}", filterCutoffDate); } String deleteCutoffDate = ""; if (doDelete) { Calendar cal = Calendar.getInstance(); cal.add(Calendar.DAY_OF_MONTH, Integer.parseInt("-" + daysBeforeDelete)); deleteCutoffDate = String.format("%04d%02d%02d%02d", cal.get(Calendar.YEAR), (cal.get(Calendar.MONTH) + 1), cal.get(Calendar.DAY_OF_MONTH), cal.get(Calendar.HOUR_OF_DAY)); LOG.info("Deleting logs from before {}", deleteCutoffDate); } long now = System.currentTimeMillis(); // Various exceptions have been popping up here. So make sure I catch them // all. try { // Patterns to recognize hour, day and incoming directories, so that they // can be processed. Pattern datePathPattern; Pattern hourPathPattern; Pattern incomingPathPattern; Pattern dataPathPattern; Pattern archivePathPattern; Pattern workingPathPattern; if (hour != null) { datePathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")"); hourPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(" + Pattern.quote(hour) + ")"); incomingPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(" + Pattern.quote(hour) + ")/([^/]+)/incoming"); dataPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(" + Pattern.quote(hour) + ")/([^/]+)/data"); archivePathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(" + Pattern.quote(hour) + ")/([^/]+)/archive"); workingPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(" + Pattern.quote(hour) + ")/([^/]+)/working/([^/]+)_(\\d+)"); } else if (date != null) { datePathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")"); hourPathPattern = Pattern .compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(\\d{2})"); incomingPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(\\d{2})/([^/]+)/incoming"); dataPathPattern = Pattern .compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(\\d{2})/([^/]+)/data"); archivePathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(\\d{2})/([^/]+)/archive"); workingPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(" + Pattern.quote(date) + ")/(\\d{2})/([^/]+)/working/([^/]+)_(\\d+)"); } else { datePathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(\\d{8})"); hourPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(\\d{8})/(\\d{2})"); incomingPathPattern = Pattern .compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(\\d{8})/(\\d{2})/([^/]+)/incoming"); dataPathPattern = Pattern.compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(\\d{8})/(\\d{2})/([^/]+)/data"); archivePathPattern = Pattern .compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(\\d{8})/(\\d{2})/([^/]+)/archive"); workingPathPattern = Pattern .compile(rootDir + "/" + Pattern.quote(dcNumber) + "/" + Pattern.quote(service) + "/" + Pattern.quote(logdir) + "/(\\d{8})/(\\d{2})/([^/]+)/working/([^/]+)_(\\d+)"); } // Do a depth first search of the directory, processing anything that // looks // interesting along the way Deque<Path> paths = new ArrayDeque<Path>(); Path rootPath = new Path(rootDir + "/" + dcNumber + "/" + service + "/" + logdir + "/"); paths.push(rootPath); while (paths.size() > 0) { Path p = paths.pop(); LOG.debug("{}", p.toString()); if (!fs.exists(p)) { continue; } FileStatus dirStatus = fs.getFileStatus(p); FileStatus[] children = fs.listStatus(p); boolean addChildren = true; boolean old = dirStatus.getModificationTime() < now - WAIT_TIME; LOG.debug(" Was last modified {}ms ago", now - dirStatus.getModificationTime()); if (!old) { LOG.debug(" Skipping, since it's not old enough."); } else if ((!rootPath.equals(p)) && (children.length == 0 || (children.length == 1 && children[0].getPath().getName().equals(READY_MARKER)))) { // old and no children? Delete! LOG.info(" Deleting empty directory {}", p.toString()); fs.delete(p, true); } else { Matcher matcher = datePathPattern.matcher(p.toUri().getPath()); if (matcher.matches()) { LOG.debug("Checking date directory"); // If this is already done, then skip it. So only process if it // doesn't exist. if (fs.exists(new Path(p, READY_MARKER)) == false) { // Check each subdirectory. If they all have ready markers, then I // guess we're ready. boolean ready = true; for (FileStatus c : children) { if (c.isDirectory() && fs.exists(new Path(c.getPath(), READY_MARKER)) == false) { ready = false; break; } } if (ready) { fs.createNewFile(new Path(p, READY_MARKER)); } } } matcher = hourPathPattern.matcher(p.toUri().getPath()); if (matcher.matches()) { LOG.debug("Checking hour directory"); // If this is already done, then skip it. So only process if it // doesn't exist. if (fs.exists(new Path(p, READY_MARKER)) == false) { // Check each subdirectory. If they all have ready markers, then I // guess we're ready. boolean ready = true; for (FileStatus c : children) { if (c.isDirectory() && fs.exists(new Path(c.getPath(), READY_MARKER)) == false) { ready = false; break; } } if (ready) { fs.createNewFile(new Path(p, READY_MARKER)); } } } // Check to see if we have to run a merge matcher = incomingPathPattern.matcher(p.toUri().getPath()); if (matcher.matches()) { LOG.debug("Checking incoming directory"); String matchDate = matcher.group(1); String matchHour = matcher.group(2); String matchComponent = matcher.group(3); String timestamp = matchDate + matchHour; if (doDelete && timestamp.compareTo(deleteCutoffDate) < 0) { LOG.info("Deleting old directory: {}", p); fs.delete(p, true); addChildren = false; } else if (doMerge) { // old, looks right, and has children? Run it! boolean hasMatchingChildren = false; boolean subdirTooYoung = false; for (FileStatus child : children) { if (!hasMatchingChildren) { FileStatus[] grandchildren = fs.listStatus(child.getPath()); for (FileStatus gc : grandchildren) { if (VALID_FILE.matcher(gc.getPath().getName()).matches()) { hasMatchingChildren = true; break; } } } if (!subdirTooYoung) { if (child.getModificationTime() >= now - WAIT_TIME) { subdirTooYoung = true; LOG.debug(" Subdir {} is too young.", child.getPath()); } } } if (!hasMatchingChildren) { LOG.debug(" No files match the expected pattern ({})", VALID_FILE.pattern()); } if (hasMatchingChildren && !subdirTooYoung) { LOG.info(" Run Merge job {} :: {} {} {} {} {}", new Object[] { p.toString(), dcNumber, service, matchDate, matchHour, matchComponent }); Properties jobProps = new Properties(); jobProps.load(new FileInputStream(mergeJobPropertiesFile)); jobProps.setProperty("jobType", "merge"); jobProps.setProperty("rootDir", rootDir); jobProps.setProperty("dcNumber", dcNumber); jobProps.setProperty("service", service); jobProps.setProperty("date", matchDate); jobProps.setProperty("hour", matchHour); jobProps.setProperty("component", matchComponent); jobProps.setProperty("user.name", userName); jobProps.setProperty("logdir", logdir); jobRunner.submit(jobProps); addChildren = false; } } } // Check to see if we need to run a filter and archive matcher = dataPathPattern.matcher(p.toUri().getPath()); if (matcher.matches()) { String matchDate = matcher.group(1); String matchHour = matcher.group(2); String matchComponent = matcher.group(3); String timestamp = matchDate + matchHour; if (doDelete && timestamp.compareTo(deleteCutoffDate) < 0) { LOG.info("Deleting old directory: {}", p); fs.delete(p, true); addChildren = false; } else if (doArchive && timestamp.compareTo(filterCutoffDate) < 0) { Properties jobProps = new Properties(); jobProps.load(new FileInputStream(filterJobPropertiesFile)); jobProps.setProperty("jobType", "filter"); jobProps.setProperty("rootDir", rootDir); jobProps.setProperty("dcNumber", dcNumber); jobProps.setProperty("service", service); jobProps.setProperty("date", matchDate); jobProps.setProperty("hour", matchHour); jobProps.setProperty("component", matchComponent); jobProps.setProperty("user.name", userName); jobProps.setProperty("logdir", logdir); // Check to see if we should just keep all or delete all here. // The filter file should be here String appPath = jobProps.getProperty("oozie.wf.application.path"); appPath = appPath.replaceFirst("\\$\\{.*?\\}", ""); Path filterFile = new Path( appPath + "/" + conf.get("filter.definition.file", service + ".yaml")); LOG.info("Filter file is {}", filterFile); if (fs.exists(filterFile)) { List<BoomFilterMapper.Filter> filters = BoomFilterMapper.loadFilters(matchComponent, fs.open(filterFile)); if (filters == null) { LOG.warn( " Got null when getting filters. Not processing. {} :: {} {} {} {} {}", new Object[] { p.toString(), dcNumber, service, matchDate, matchHour, matchComponent }); } else if (filters.size() == 0) { LOG.warn(" Got no filters. Not processing. {} :: {} {} {} {} {}", new Object[] { p.toString(), dcNumber, service, matchDate, matchHour, matchComponent }); } else if (filters.size() == 1 && filters.get(0) instanceof BoomFilterMapper.KeepAllFilter) { LOG.info(" Keeping everything. {} :: {} {} {} {} {}", new Object[] { p.toString(), dcNumber, service, matchDate, matchHour, matchComponent }); // Move files from data to archive // delete it all! String destination = rootDir + "/" + dcNumber + "/" + service + "/" + logdir + "/" + matchDate + "/" + matchHour + "/" + matchComponent + "/archive/"; PathInfo pathInfo = new PathInfo(); pathInfo.setDcNumber(dcNumber); pathInfo.setService(service); pathInfo.setLogdir(logdir); pathInfo.setDate(matchDate); pathInfo.setHour(matchHour); pathInfo.setComponent(matchComponent); try { lockUtil.acquireWriteLock(lockUtil.getLockPath(pathInfo)); fs.mkdirs(new Path(destination)); for (FileStatus f : fs.listStatus(p)) { fs.rename(f.getPath(), new Path(destination)); } } finally { lockUtil.releaseWriteLock(lockUtil.getLockPath(pathInfo)); } } else if (filters.size() == 1 && filters.get(0) instanceof BoomFilterMapper.DropAllFilter) { LOG.info(" Dropping everything. {} :: {} {} {} {} {}", new Object[] { p.toString(), dcNumber, service, matchDate, matchHour, matchComponent }); PathInfo pathInfo = new PathInfo(); pathInfo.setDcNumber(dcNumber); pathInfo.setService(service); pathInfo.setLogdir(logdir); pathInfo.setDate(matchDate); pathInfo.setHour(matchHour); pathInfo.setComponent(matchComponent); try { lockUtil.acquireWriteLock(lockUtil.getLockPath(pathInfo)); fs.delete(p, true); } finally { lockUtil.releaseWriteLock(lockUtil.getLockPath(pathInfo)); } } else { LOG.info(" Run Filter/Archive job {} :: {} {} {} {} {}", new Object[] { p.toString(), dcNumber, service, matchDate, matchHour, matchComponent }); jobRunner.submit(jobProps); } } else { LOG.warn("Skipping filter job, since no filter file exists"); } addChildren = false; } } matcher = archivePathPattern.matcher(p.toUri().getPath()); if (matcher.matches()) { String matchDate = matcher.group(1); String matchHour = matcher.group(2); String timestamp = matchDate + matchHour; if (doDelete && timestamp.compareTo(deleteCutoffDate) < 0) { LOG.info("Deleting old directory: {}", p); fs.delete(p, true); addChildren = false; } } matcher = workingPathPattern.matcher(p.toUri().getPath()); if (matcher.matches()) { LOG.info(" Matches working pattern ({})", p); if (resetOrphanedJobs) { String matchDate = matcher.group(1); String matchHour = matcher.group(2); String matchComponent = matcher.group(3); // Move everything from working/xxx/incoming/ to incoming/ PathInfo lockPathInfo = new PathInfo(logdir, rootDir + "/" + dcNumber + "/" + service + "/" + logdir + "/" + matchDate + "/" + matchHour + "/" + matchComponent); lockUtil.acquireWriteLock(lockUtil.getLockPath(lockPathInfo)); FileStatus[] fileStatuses = fs.listStatus(new Path(p.toUri().getPath() + "/incoming/")); if (fileStatuses != null) { for (FileStatus fileStatus : fileStatuses) { Path toPath = new Path( fileStatus.getPath().getParent().getParent().getParent().getParent(), "incoming/" + fileStatus.getPath().getName()); LOG.info(" Moving data from {} to {}", fileStatus.getPath(), toPath); LOG.info(" mkdir {}", toPath); fs.mkdirs(toPath); Path fromDir = new Path(p.toUri().getPath(), "incoming/" + fileStatus.getPath().getName()); LOG.info(" moving from {}", fromDir); FileStatus[] files = fs.listStatus(fromDir); if (files == null || files.length == 0) { LOG.info(" Nothing to move from {}", fromDir); } else { for (FileStatus f : files) { LOG.info(" rename {} {}", f.getPath(), new Path(toPath, f.getPath().getName())); fs.rename(f.getPath(), new Path(toPath, f.getPath().getName())); } } LOG.info(" rm {}", fileStatus.getPath()); fs.delete(fileStatus.getPath(), true); } lockUtil.releaseWriteLock(lockUtil.getLockPath(lockPathInfo)); fs.delete(new Path(p.toUri().getPath()), true); } } addChildren = false; } } // Add any children which are directories to the stack. if (addChildren) { for (int i = children.length - 1; i >= 0; i--) { FileStatus child = children[i]; if (child.isDirectory()) { paths.push(child.getPath()); } } } } // Since we may have deleted a bunch of directories, delete any unused // locks // from ZooKeeper. { LOG.info("Checking for unused locks in ZooKeeper"); String scanPath = rootDir + "/" + dcNumber + "/" + service + "/" + logdir; if (date != null) { scanPath += "/" + date; if (hour != null) { scanPath += "/" + hour; } } List<LockInfo> lockInfo = lockUtil.scan(scanPath); for (LockInfo li : lockInfo) { // Check if the lock path still exists in HDFS. If it doesn't, then // delete it from ZooKeeper. String path = li.getPath(); String hdfsPath = path.substring(LockUtil.ROOT.length()); if (!fs.exists(new Path(hdfsPath))) { ZooKeeper zk = lockUtil.getZkClient(); while (!path.equals(LockUtil.ROOT)) { try { zk.delete(path, -1); } catch (KeeperException.NotEmptyException e) { // That's fine. just stop trying then. break; } catch (Exception e) { LOG.error("Caught exception trying to delete from ZooKeeper.", e); break; } LOG.info("Deleted from ZooKeeper: {}", path); path = path.substring(0, path.lastIndexOf('/')); } } } } // Now that we're done, wait for the Oozie Runner to stop, and print the // results. LOG.info("Waiting for Oozie jobs to complete."); jobRunner.shutdown(); jobRunnerThread.join(); LOG.info("Job Stats : Started={} Succeeded={} failed={} errors={}", new Object[] { jobRunner.getStarted(), jobRunner.getSucceeded(), jobRunner.getFailed(), jobRunner.getErrors() }); lockUtil.close(); } catch (Exception e) { LOG.error("Unexpected exception caught.", e); return 1; } return 0; }
From source file:com.ceph.rados.fs.RadosFileSystemStore.java
License:Apache License
private String pathToKey(Path path) { if (!path.isAbsolute()) { throw new IllegalArgumentException("Path must be absolute: " + path); }/*ww w .j av a 2s .c o m*/ return path.toUri().getPath(); }
From source file:com.cfets.door.yarn.jboss.JBossClient.java
License:Apache License
/** * Main run function for the client/*from w w w. ja v a2 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()); } } YarnClientApplication app = yarnClient.createApplication(); GetNewApplicationResponse appResponse = app.getNewApplicationResponse(); int maxMem = appResponse.getMaximumResourceCapability().getMemory(); LOG.info("Max mem capabililty 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; } ApplicationSubmissionContext appContext = app.getApplicationSubmissionContext(); ApplicationId appId = appContext.getApplicationId(); appContext.setApplicationName(appName); ContainerLaunchContext amContainer = Records.newRecord(ContainerLaunchContext.class); Map<String, LocalResource> localResources = new HashMap<String, LocalResource>(); LOG.info("Copy App Master jar from local filesystem and add to local environment"); FileSystem fs = FileSystem.get(conf); Path src = new Path(appJar); String pathSuffix = appName + File.separator + appId.getId() + File.separator + JBossConstants.JBOSS_ON_YARN_APP; Path dst = new Path(fs.getHomeDirectory(), pathSuffix); jbossAppUri = dst.toUri().toString(); fs.copyFromLocalFile(false, true, src, dst); FileStatus destStatus = fs.getFileStatus(dst); LocalResource amJarRsrc = Records.newRecord(LocalResource.class); amJarRsrc.setType(LocalResourceType.FILE); amJarRsrc.setVisibility(LocalResourceVisibility.APPLICATION); amJarRsrc.setResource(ConverterUtils.getYarnUrlFromPath(dst)); amJarRsrc.setTimestamp(destStatus.getModificationTime()); amJarRsrc.setSize(destStatus.getLen()); localResources.put(JBossConstants.JBOSS_ON_YARN_APP, amJarRsrc); 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); } amContainer.setLocalResources(localResources); LOG.info("Set the environment for the application master"); Map<String, String> env = new HashMap<String, String>(); StringBuilder classPathEnv = new StringBuilder(Environment.CLASSPATH.$()).append(File.pathSeparatorChar) .append("./*"); for (String c : conf.getStrings(YarnConfiguration.YARN_APPLICATION_CLASSPATH, YarnConfiguration.DEFAULT_YARN_APPLICATION_CLASSPATH)) { classPathEnv.append(File.pathSeparatorChar); classPathEnv.append(c.trim()); } classPathEnv.append(File.pathSeparatorChar).append("./log4j.properties"); if (conf.getBoolean(YarnConfiguration.IS_MINI_YARN_CLUSTER, false)) { classPathEnv.append(':'); classPathEnv.append(System.getProperty("java.class.path")); } env.put("CLASSPATH", classPathEnv.toString()); amContainer.setEnvironment(env); Vector<CharSequence> vargs = new Vector<CharSequence>(30); LOG.info("Setting up app master command"); vargs.add(Environment.JAVA_HOME.$() + "/bin/java"); vargs.add("-Xmx" + amMemory + "m"); vargs.add(appMasterMainClass); vargs.add("--container_memory " + String.valueOf(containerMemory)); vargs.add("--num_containers " + String.valueOf(numContainers)); vargs.add("--priority " + String.valueOf(shellCmdPriority)); vargs.add("--admin_user " + adminUser); vargs.add("--admin_password " + adminPassword); vargs.add("--jar " + jbossAppUri); if (debugFlag) { vargs.add("--debug"); } vargs.add("1>" + JBossConstants.JBOSS_CONTAINER_LOG_DIR + "/JBossApplicationMaster.stdout"); vargs.add("2>" + JBossConstants.JBOSS_CONTAINER_LOG_DIR + "/JBossApplicationMaster.stderr"); 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); Resource capability = Records.newRecord(Resource.class); capability.setMemory(amMemory); appContext.setResource(capability); appContext.setAMContainerSpec(amContainer); Priority pri = Records.newRecord(Priority.class); pri.setPriority(amPriority); appContext.setPriority(pri); appContext.setQueue(amQueue); LOG.info("Submitting the application to ASM"); yarnClient.submitApplication(appContext); return monitorApplication(appId); }
From source file:com.cg.mapreduce.fpgrowth.mahout.fpm.FPGrowthDriver.java
License:Apache License
private static void runFPGrowth(Parameters params) throws IOException { log.info("Starting Sequential FPGrowth"); int maxHeapSize = Integer.valueOf(params.get("maxHeapSize", "50")); int minSupport = Integer.valueOf(params.get("minSupport", "3")); Path output = new Path(params.get("output", "output.txt")); Path input = new Path(params.get("input")); Configuration conf = new Configuration(); FileSystem fs = FileSystem.get(output.toUri(), conf); Charset encoding = Charset.forName(params.get("encoding")); String pattern = params.get("splitPattern", PFPGrowth.SPLITTER.toString()); SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, output, Text.class, TopKStringPatterns.class); FSDataInputStream inputStream = null; FSDataInputStream inputStreamAgain = null; Collection<String> features = Sets.newHashSet(); if ("true".equals(params.get(PFPGrowth.USE_FPG2))) { com.cg.mapreduce.fpgrowth.mahout.fpm.fpgrowth2.FPGrowthObj<String> fp = new com.cg.mapreduce.fpgrowth.mahout.fpm.fpgrowth2.FPGrowthObj<String>(); try {/*w w w . jav a 2 s.c o m*/ inputStream = fs.open(input); inputStreamAgain = fs.open(input); fp.generateTopKFrequentPatterns( new StringRecordIterator(new FileLineIterable(inputStream, encoding, false), pattern), fp.generateFList(new StringRecordIterator( new FileLineIterable(inputStreamAgain, encoding, false), pattern), minSupport), minSupport, maxHeapSize, features, new StringOutputConverter( new SequenceFileOutputCollector<Text, TopKStringPatterns>(writer)), new ContextStatusUpdater(null)); } finally { Closeables.close(writer, false); Closeables.close(inputStream, true); Closeables.close(inputStreamAgain, true); } } else { FPGrowth<String> fp = new FPGrowth<String>(); inputStream = fs.open(input); inputStreamAgain = fs.open(input); try { fp.generateTopKFrequentPatterns( new StringRecordIterator(new FileLineIterable(inputStream, encoding, false), pattern), fp.generateFList(new StringRecordIterator( new FileLineIterable(inputStreamAgain, encoding, false), pattern), minSupport), minSupport, maxHeapSize, features, new StringOutputConverter( new SequenceFileOutputCollector<Text, TopKStringPatterns>(writer)), new ContextStatusUpdater(null)); } finally { Closeables.close(writer, false); Closeables.close(inputStream, true); Closeables.close(inputStreamAgain, true); } } List<Pair<String, TopKStringPatterns>> frequentPatterns = FPGrowth.readFrequentPattern(conf, output); for (Pair<String, TopKStringPatterns> entry : frequentPatterns) { log.info("Dumping Patterns for Feature: {} \n{}", entry.getFirst(), entry.getSecond()); } }
From source file:com.cg.mapreduce.fpgrowth.mahout.fpm.PFPGrowth.java
License:Apache License
/** * Serializes the fList and returns the string representation of the List *//*from ww w . jav a 2 s .c o m*/ public static void saveFList(Iterable<Pair<String, Long>> flist, Parameters params, Configuration conf) throws IOException { Path flistPath = new Path(params.get(OUTPUT), F_LIST); FileSystem fs = FileSystem.get(flistPath.toUri(), conf); flistPath = fs.makeQualified(flistPath); HadoopUtil.delete(conf, flistPath); SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, flistPath, Text.class, LongWritable.class); try { for (Pair<String, Long> pair : flist) { writer.append(new Text(pair.getFirst()), new LongWritable(pair.getSecond())); } } finally { writer.close(); } DistributedCache.addCacheFile(flistPath.toUri(), conf); }
From source file:com.cg.mapreduce.fpgrowth.mahout.fpm.PFPGrowth.java
License:Apache License
/** * Read the Frequent Patterns generated from Text * // w w w .jav a 2s . c o m * @return List of TopK patterns for each string frequent feature */ public static List<Pair<String, TopKStringPatterns>> readFrequentPattern(Parameters params) throws IOException { Configuration conf = new Configuration(); Path frequentPatternsPath = new Path(params.get(OUTPUT), FREQUENT_PATTERNS); FileSystem fs = FileSystem.get(frequentPatternsPath.toUri(), conf); FileStatus[] outputFiles = fs.globStatus(new Path(frequentPatternsPath, FILE_PATTERN)); List<Pair<String, TopKStringPatterns>> ret = Lists.newArrayList(); for (FileStatus fileStatus : outputFiles) { ret.addAll(FPGrowth.readFrequentPattern(conf, fileStatus.getPath())); } return ret; }
From source file:com.cg.mapreduce.myfpgrowth.PFPGrowth.java
License:Apache License
/** * Serializes the fList and returns the string representation of the List *//*from w ww . j av a 2 s. c o m*/ public static void saveFList(List<Pair<String, Long>> fList, Parameters params, Configuration conf) throws IOException { Path flistPath = new Path(params.get(OUTPUT) + "/oldlist", F_LIST); FileSystem fs = FileSystem.get(flistPath.toUri(), conf); flistPath = fs.makeQualified(flistPath); HadoopUtil.delete(conf, flistPath); SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, flistPath, Text.class, Pair.class); // set param to control group size in MR jobs int numGroups = params.getInt(NUM_GROUPS, NUM_GROUPS_DEFAULT); int maxPerGroup = fList.size() / numGroups; if (fList.size() % numGroups != 0) { maxPerGroup++; } params.set(MAX_PER_GROUP, Integer.toString(maxPerGroup)); try { int group = 0; int count = 0; for (Pair<String, Long> pair : fList) { if (count == maxPerGroup) { group++; count = 0; } writer.append(new Text(pair.getFirst()), new Pair<Integer, Long>(group, pair.getSecond())); //writer.append(new Text(pair.getFirst()), new LongWritable(pair.getSecond())); } } finally { writer.close(); } DistributedCache.addCacheFile(flistPath.toUri(), conf); }