YARN-2556. Tool to measure the performance of the timeline server (Chang Li via sjlee)

(cherry picked from commit 58590fef49)
This commit is contained in:
Sangjin Lee 2015-10-16 16:53:20 -07:00
parent 1ba73169c7
commit 180efe6677
8 changed files with 917 additions and 20 deletions

View File

@ -0,0 +1,53 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapreduce;
import java.io.IOException;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.mapreduce.jobhistory.JobHistoryParser;
import org.apache.hadoop.mapreduce.jobhistory.JobHistoryParser.JobInfo;
class JobHistoryFileParser {
private static final Log LOG = LogFactory.getLog(JobHistoryFileParser.class);
private final FileSystem fs;
public JobHistoryFileParser(FileSystem fs) {
LOG.info("JobHistoryFileParser created with " + fs);
this.fs = fs;
}
public JobInfo parseHistoryFile(Path path) throws IOException {
LOG.info("parsing job history file " + path);
JobHistoryParser parser = new JobHistoryParser(fs, path);
return parser.parse();
}
public Configuration parseConfiguration(Path path) throws IOException {
LOG.info("parsing job configuration file " + path);
Configuration conf = new Configuration(false);
conf.addResource(fs.open(path));
return conf;
}
}

View File

@ -0,0 +1,196 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapreduce;
import java.io.IOException;
import java.util.Collection;
import java.util.HashMap;
import java.util.Map;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.LocatedFileStatus;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.fs.RemoteIterator;
import org.apache.hadoop.mapreduce.Mapper.Context;
class JobHistoryFileReplayHelper {
private static final Log LOG =
LogFactory.getLog(JobHistoryFileReplayHelper.class);
static final String PROCESSING_PATH = "processing path";
static final String REPLAY_MODE = "replay mode";
static final int WRITE_ALL_AT_ONCE = 1;
static final int WRITE_PER_ENTITY = 2;
static final int REPLAY_MODE_DEFAULT = WRITE_ALL_AT_ONCE;
private static Pattern JOB_ID_PARSER =
Pattern.compile("^(job_[0-9]+_([0-9]+)).*");
private enum FileType { JOB_HISTORY_FILE, JOB_CONF_FILE, UNKNOWN };
JobHistoryFileParser parser;
int replayMode;
Collection<JobFiles> jobFiles;
JobHistoryFileReplayHelper(Context context) throws IOException {
Configuration conf = context.getConfiguration();
int taskId = context.getTaskAttemptID().getTaskID().getId();
int size = conf.getInt(MRJobConfig.NUM_MAPS,
TimelineServicePerformance.NUM_MAPS_DEFAULT);
replayMode = conf.getInt(JobHistoryFileReplayHelper.REPLAY_MODE,
JobHistoryFileReplayHelper.REPLAY_MODE_DEFAULT);
String processingDir =
conf.get(JobHistoryFileReplayHelper.PROCESSING_PATH);
Path processingPath = new Path(processingDir);
FileSystem processingFs = processingPath.getFileSystem(conf);
parser = new JobHistoryFileParser(processingFs);
jobFiles = selectJobFiles(processingFs, processingPath, taskId, size);
}
public int getReplayMode() {
return replayMode;
}
public Collection<JobFiles> getJobFiles() {
return jobFiles;
}
public JobHistoryFileParser getParser() {
return parser;
}
public static class JobFiles {
private final String jobId;
private Path jobHistoryFilePath;
private Path jobConfFilePath;
public JobFiles(String jobId) {
this.jobId = jobId;
}
public String getJobId() {
return jobId;
}
public Path getJobHistoryFilePath() {
return jobHistoryFilePath;
}
public void setJobHistoryFilePath(Path jobHistoryFilePath) {
this.jobHistoryFilePath = jobHistoryFilePath;
}
public Path getJobConfFilePath() {
return jobConfFilePath;
}
public void setJobConfFilePath(Path jobConfFilePath) {
this.jobConfFilePath = jobConfFilePath;
}
@Override
public int hashCode() {
return jobId.hashCode();
}
@Override
public boolean equals(Object obj) {
if (this == obj) {
return true;
}
if (obj == null) {
return false;
}
if (getClass() != obj.getClass()) {
return false;
}
JobFiles other = (JobFiles) obj;
return jobId.equals(other.jobId);
}
}
private Collection<JobFiles> selectJobFiles(FileSystem fs,
Path processingRoot, int i, int size) throws IOException {
Map<String, JobFiles> jobs = new HashMap<>();
RemoteIterator<LocatedFileStatus> it = fs.listFiles(processingRoot, true);
while (it.hasNext()) {
LocatedFileStatus status = it.next();
Path path = status.getPath();
String fileName = path.getName();
Matcher m = JOB_ID_PARSER.matcher(fileName);
if (!m.matches()) {
continue;
}
String jobId = m.group(1);
int lastId = Integer.parseInt(m.group(2));
int mod = lastId % size;
if (mod != i) {
continue;
}
LOG.info("this mapper will process file " + fileName);
// it's mine
JobFiles jobFiles = jobs.get(jobId);
if (jobFiles == null) {
jobFiles = new JobFiles(jobId);
jobs.put(jobId, jobFiles);
}
setFilePath(fileName, path, jobFiles);
}
return jobs.values();
}
private void setFilePath(String fileName, Path path,
JobFiles jobFiles) {
// determine if we're dealing with a job history file or a job conf file
FileType type = getFileType(fileName);
switch (type) {
case JOB_HISTORY_FILE:
if (jobFiles.getJobHistoryFilePath() == null) {
jobFiles.setJobHistoryFilePath(path);
} else {
LOG.warn("we already have the job history file " +
jobFiles.getJobHistoryFilePath() + ": skipping " + path);
}
break;
case JOB_CONF_FILE:
if (jobFiles.getJobConfFilePath() == null) {
jobFiles.setJobConfFilePath(path);
} else {
LOG.warn("we already have the job conf file " +
jobFiles.getJobConfFilePath() + ": skipping " + path);
}
break;
case UNKNOWN:
LOG.warn("unknown type: " + path);
}
}
private FileType getFileType(String fileName) {
if (fileName.endsWith(".jhist")) {
return FileType.JOB_HISTORY_FILE;
}
if (fileName.endsWith("_conf.xml")) {
return FileType.JOB_CONF_FILE;
}
return FileType.UNKNOWN;
}
}

View File

@ -0,0 +1,158 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapreduce;
import java.io.IOException;
import java.util.Collection;
import java.util.HashMap;
import java.util.Map;
import java.util.Set;
import java.util.concurrent.TimeUnit;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.LocatedFileStatus;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.fs.RemoteIterator;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapreduce.Mapper.Context;
import org.apache.hadoop.mapreduce.TimelineServicePerformance.PerfCounters;
import org.apache.hadoop.mapreduce.JobHistoryFileReplayHelper;
import org.apache.hadoop.mapreduce.JobHistoryFileReplayHelper.JobFiles;
import org.apache.hadoop.mapreduce.MRJobConfig;
import org.apache.hadoop.mapreduce.TypeConverter;
import org.apache.hadoop.mapreduce.jobhistory.JobHistoryParser.JobInfo;
import org.apache.hadoop.mapreduce.v2.api.records.JobId;
import org.apache.hadoop.security.UserGroupInformation;
import org.apache.hadoop.yarn.api.records.ApplicationId;
import org.apache.hadoop.yarn.api.records.timeline.TimelineEntities;
import org.apache.hadoop.yarn.api.records.timeline.TimelineEntity;
import org.apache.hadoop.yarn.client.api.TimelineClient;
import org.apache.hadoop.yarn.client.api.impl.TimelineClientImpl;
import org.apache.hadoop.yarn.exceptions.YarnException;
/**
* Mapper for TimelineServicePerformanceV1 that replays job history files to the
* timeline service.
*
*/
class JobHistoryFileReplayMapperV1 extends
org.apache.hadoop.mapreduce.
Mapper<IntWritable,IntWritable,Writable,Writable> {
private static final Log LOG =
LogFactory.getLog(JobHistoryFileReplayMapperV1.class);
public void map(IntWritable key, IntWritable val, Context context) throws IOException {
// collect the apps it needs to process
TimelineClient tlc = new TimelineClientImpl();
TimelineEntityConverterV1 converter = new TimelineEntityConverterV1();
JobHistoryFileReplayHelper helper = new JobHistoryFileReplayHelper(context);
int replayMode = helper.getReplayMode();
Collection<JobFiles> jobs =
helper.getJobFiles();
JobHistoryFileParser parser = helper.getParser();
if (jobs.isEmpty()) {
LOG.info(context.getTaskAttemptID().getTaskID() +
" will process no jobs");
} else {
LOG.info(context.getTaskAttemptID().getTaskID() + " will process " +
jobs.size() + " jobs");
}
for (JobFiles job: jobs) {
// process each job
String jobIdStr = job.getJobId();
LOG.info("processing " + jobIdStr + "...");
JobId jobId = TypeConverter.toYarn(JobID.forName(jobIdStr));
ApplicationId appId = jobId.getAppId();
try {
// parse the job info and configuration
Path historyFilePath = job.getJobHistoryFilePath();
Path confFilePath = job.getJobConfFilePath();
if ((historyFilePath == null) || (confFilePath == null)) {
continue;
}
JobInfo jobInfo =
parser.parseHistoryFile(historyFilePath);
Configuration jobConf =
parser.parseConfiguration(confFilePath);
LOG.info("parsed the job history file and the configuration file for job "
+ jobIdStr);
// create entities from job history and write them
long totalTime = 0;
Set<TimelineEntity> entitySet =
converter.createTimelineEntities(jobInfo, jobConf);
LOG.info("converted them into timeline entities for job " + jobIdStr);
// use the current user for this purpose
UserGroupInformation ugi = UserGroupInformation.getCurrentUser();
long startWrite = System.nanoTime();
try {
switch (replayMode) {
case JobHistoryFileReplayHelper.WRITE_ALL_AT_ONCE:
writeAllEntities(tlc, entitySet, ugi);
break;
case JobHistoryFileReplayHelper.WRITE_PER_ENTITY:
writePerEntity(tlc, entitySet, ugi);
break;
default:
break;
}
} catch (Exception e) {
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_FAILURES).
increment(1);
LOG.error("writing to the timeline service failed", e);
}
long endWrite = System.nanoTime();
totalTime += TimeUnit.NANOSECONDS.toMillis(endWrite-startWrite);
int numEntities = entitySet.size();
LOG.info("wrote " + numEntities + " entities in " + totalTime + " ms");
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_TIME).
increment(totalTime);
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_COUNTER).
increment(numEntities);
} finally {
context.progress(); // move it along
}
}
}
private void writeAllEntities(TimelineClient tlc,
Set<TimelineEntity> entitySet, UserGroupInformation ugi)
throws IOException, YarnException {
tlc.putEntities((TimelineEntity[])entitySet.toArray());
}
private void writePerEntity(TimelineClient tlc,
Set<TimelineEntity> entitySet, UserGroupInformation ugi)
throws IOException, YarnException {
for (TimelineEntity entity : entitySet) {
tlc.putEntities(entity);
LOG.info("wrote entity " + entity.getEntityId());
}
}
}

View File

@ -0,0 +1,120 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapreduce;
import java.io.IOException;
import java.util.Random;
import java.util.concurrent.TimeUnit;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapreduce.Mapper.Context;
import org.apache.hadoop.mapreduce.TimelineServicePerformance.PerfCounters;
import org.apache.hadoop.mapreduce.TaskAttemptID;
import org.apache.hadoop.security.UserGroupInformation;
import org.apache.hadoop.yarn.api.records.ApplicationId;
import org.apache.hadoop.yarn.api.records.timeline.TimelineEntities;
import org.apache.hadoop.yarn.api.records.timeline.TimelineEntity;
import org.apache.hadoop.yarn.api.records.timeline.TimelineEvent;
import org.apache.hadoop.yarn.client.api.TimelineClient;
import org.apache.hadoop.yarn.client.api.impl.TimelineClientImpl;
import org.apache.hadoop.yarn.conf.YarnConfiguration;
/**
* Adds simple entities with random string payload, events, metrics, and
* configuration.
*/
class SimpleEntityWriterV1 extends
org.apache.hadoop.mapreduce.Mapper<IntWritable,IntWritable,Writable,Writable> {
private static final Log LOG = LogFactory.getLog(SimpleEntityWriterV1.class);
// constants for mtype = 1
static final String KBS_SENT = "kbs sent";
static final int KBS_SENT_DEFAULT = 1;
static final String TEST_TIMES = "testtimes";
static final int TEST_TIMES_DEFAULT = 100;
static final String TIMELINE_SERVICE_PERFORMANCE_RUN_ID =
"timeline.server.performance.run.id";
/**
* To ensure that the compression really gets exercised, generate a
* random alphanumeric fixed length payload
*/
private static char[] ALPHA_NUMS = new char[] { 'a', 'b', 'c', 'd', 'e', 'f',
'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r',
's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 'A', 'B', 'C', 'D',
'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',
'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', '1', '2',
'3', '4', '5', '6', '7', '8', '9', '0', ' ' };
public void map(IntWritable key, IntWritable val, Context context) throws IOException {
TimelineClient tlc = new TimelineClientImpl();
Configuration conf = context.getConfiguration();
final int kbs = conf.getInt(KBS_SENT, KBS_SENT_DEFAULT);
long totalTime = 0;
final int testtimes = conf.getInt(TEST_TIMES, TEST_TIMES_DEFAULT);
final Random rand = new Random();
final TaskAttemptID taskAttemptId = context.getTaskAttemptID();
final char[] payLoad = new char[kbs * 1024];
for (int i = 0; i < testtimes; i++) {
// Generate a fixed length random payload
for (int xx = 0; xx < kbs * 1024; xx++) {
int alphaNumIdx =
rand.nextInt(ALPHA_NUMS.length);
payLoad[xx] = ALPHA_NUMS[alphaNumIdx];
}
String entId = taskAttemptId + "_" + Integer.toString(i);
final TimelineEntity entity = new TimelineEntity();
entity.setEntityId(entId);
entity.setEntityType("FOO_ATTEMPT");
entity.addOtherInfo("PERF_TEST", payLoad);
// add an event
TimelineEvent event = new TimelineEvent();
event.setTimestamp(System.currentTimeMillis());
event.setEventType("foo_event");
entity.addEvent(event);
// use the current user for this purpose
UserGroupInformation ugi = UserGroupInformation.getCurrentUser();
long startWrite = System.nanoTime();
try {
tlc.putEntities(entity);
} catch (Exception e) {
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_FAILURES).
increment(1);
LOG.error("writing to the timeline service failed", e);
}
long endWrite = System.nanoTime();
totalTime += TimeUnit.NANOSECONDS.toMillis(endWrite-startWrite);
}
LOG.info("wrote " + testtimes + " entities (" + kbs*testtimes +
" kB) in " + totalTime + " ms");
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_TIME).
increment(totalTime);
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_COUNTER).
increment(testtimes);
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_KBS).
increment(kbs*testtimes);
}
}

View File

@ -0,0 +1,167 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapreduce;
import java.util.HashSet;
import java.util.Map;
import java.util.Set;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.mapreduce.Counter;
import org.apache.hadoop.mapreduce.CounterGroup;
import org.apache.hadoop.mapreduce.Counters;
import org.apache.hadoop.mapreduce.TaskAttemptID;
import org.apache.hadoop.mapreduce.TaskID;
import org.apache.hadoop.mapreduce.jobhistory.JobHistoryParser.JobInfo;
import org.apache.hadoop.mapreduce.jobhistory.JobHistoryParser.TaskAttemptInfo;
import org.apache.hadoop.mapreduce.jobhistory.JobHistoryParser.TaskInfo;
import org.apache.hadoop.yarn.api.records.timeline.TimelineEntity;
class TimelineEntityConverterV1 {
private static final Log LOG =
LogFactory.getLog(TimelineEntityConverterV1.class);
static final String JOB = "MAPREDUCE_JOB";
static final String TASK = "MAPREDUCE_TASK";
static final String TASK_ATTEMPT = "MAPREDUCE_TASK_ATTEMPT";
/**
* Creates job, task, and task attempt entities based on the job history info
* and configuration.
*
* Note: currently these are plan timeline entities created for mapreduce
* types. These are not meant to be the complete and accurate entity set-up
* for mapreduce jobs. We do not leverage hierarchical timeline entities. If
* we create canonical mapreduce hierarchical timeline entities with proper
* parent-child relationship, we could modify this to use that instead.
*
* Note that we also do not add info to the YARN application entity, which
* would be needed for aggregation.
*/
public Set<TimelineEntity> createTimelineEntities(JobInfo jobInfo,
Configuration conf) {
Set<TimelineEntity> entities = new HashSet<>();
// create the job entity
TimelineEntity job = createJobEntity(jobInfo, conf);
entities.add(job);
// create the task and task attempt entities
Set<TimelineEntity> tasksAndAttempts =
createTaskAndTaskAttemptEntities(jobInfo);
entities.addAll(tasksAndAttempts);
return entities;
}
private TimelineEntity createJobEntity(JobInfo jobInfo, Configuration conf) {
TimelineEntity job = new TimelineEntity();
job.setEntityType(JOB);
job.setEntityId(jobInfo.getJobId().toString());
job.setStartTime(jobInfo.getSubmitTime());
job.addPrimaryFilter("JOBNAME", jobInfo.getJobname());
job.addPrimaryFilter("USERNAME", jobInfo.getUsername());
job.addOtherInfo("JOB_QUEUE_NAME", jobInfo.getJobQueueName());
job.addOtherInfo("SUBMIT_TIME", jobInfo.getSubmitTime());
job.addOtherInfo("LAUNCH_TIME", jobInfo.getLaunchTime());
job.addOtherInfo("FINISH_TIME", jobInfo.getFinishTime());
job.addOtherInfo("JOB_STATUS", jobInfo.getJobStatus());
job.addOtherInfo("PRIORITY", jobInfo.getPriority());
job.addOtherInfo("TOTAL_MAPS", jobInfo.getTotalMaps());
job.addOtherInfo("TOTAL_REDUCES", jobInfo.getTotalReduces());
job.addOtherInfo("UBERIZED", jobInfo.getUberized());
job.addOtherInfo("ERROR_INFO", jobInfo.getErrorInfo());
LOG.info("converted job " + jobInfo.getJobId() + " to a timeline entity");
return job;
}
private Set<TimelineEntity> createTaskAndTaskAttemptEntities(JobInfo jobInfo) {
Set<TimelineEntity> entities = new HashSet<>();
Map<TaskID,TaskInfo> taskInfoMap = jobInfo.getAllTasks();
LOG.info("job " + jobInfo.getJobId()+ " has " + taskInfoMap.size() +
" tasks");
for (TaskInfo taskInfo: taskInfoMap.values()) {
TimelineEntity task = createTaskEntity(taskInfo);
entities.add(task);
// add the task attempts from this task
Set<TimelineEntity> taskAttempts = createTaskAttemptEntities(taskInfo);
entities.addAll(taskAttempts);
}
return entities;
}
private TimelineEntity createTaskEntity(TaskInfo taskInfo) {
TimelineEntity task = new TimelineEntity();
task.setEntityType(TASK);
task.setEntityId(taskInfo.getTaskId().toString());
task.setStartTime(taskInfo.getStartTime());
task.addOtherInfo("START_TIME", taskInfo.getStartTime());
task.addOtherInfo("FINISH_TIME", taskInfo.getFinishTime());
task.addOtherInfo("TASK_TYPE", taskInfo.getTaskType());
task.addOtherInfo("TASK_STATUS", taskInfo.getTaskStatus());
task.addOtherInfo("ERROR_INFO", taskInfo.getError());
LOG.info("converted task " + taskInfo.getTaskId() +
" to a timeline entity");
return task;
}
private Set<TimelineEntity> createTaskAttemptEntities(TaskInfo taskInfo) {
Set<TimelineEntity> taskAttempts = new HashSet<TimelineEntity>();
Map<TaskAttemptID,TaskAttemptInfo> taskAttemptInfoMap =
taskInfo.getAllTaskAttempts();
LOG.info("task " + taskInfo.getTaskId() + " has " +
taskAttemptInfoMap.size() + " task attempts");
for (TaskAttemptInfo taskAttemptInfo: taskAttemptInfoMap.values()) {
TimelineEntity taskAttempt = createTaskAttemptEntity(taskAttemptInfo);
taskAttempts.add(taskAttempt);
}
return taskAttempts;
}
private TimelineEntity createTaskAttemptEntity(TaskAttemptInfo taskAttemptInfo) {
TimelineEntity taskAttempt = new TimelineEntity();
taskAttempt.setEntityType(TASK_ATTEMPT);
taskAttempt.setEntityId(taskAttemptInfo.getAttemptId().toString());
taskAttempt.setStartTime(taskAttemptInfo.getStartTime());
taskAttempt.addOtherInfo("START_TIME", taskAttemptInfo.getStartTime());
taskAttempt.addOtherInfo("FINISH_TIME", taskAttemptInfo.getFinishTime());
taskAttempt.addOtherInfo("MAP_FINISH_TIME",
taskAttemptInfo.getMapFinishTime());
taskAttempt.addOtherInfo("SHUFFLE_FINISH_TIME",
taskAttemptInfo.getShuffleFinishTime());
taskAttempt.addOtherInfo("SORT_FINISH_TIME",
taskAttemptInfo.getSortFinishTime());
taskAttempt.addOtherInfo("TASK_STATUS", taskAttemptInfo.getTaskStatus());
taskAttempt.addOtherInfo("STATE", taskAttemptInfo.getState());
taskAttempt.addOtherInfo("ERROR", taskAttemptInfo.getError());
taskAttempt.addOtherInfo("CONTAINER_ID",
taskAttemptInfo.getContainerId().toString());
LOG.info("converted task attempt " + taskAttemptInfo.getAttemptId() +
" to a timeline entity");
return taskAttempt;
}
}

View File

@ -0,0 +1,197 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapreduce;
import java.io.IOException;
import java.util.Date;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.MRJobConfig;
import org.apache.hadoop.mapreduce.SleepJob.SleepInputFormat;
import org.apache.hadoop.mapreduce.lib.output.NullOutputFormat;
import org.apache.hadoop.util.GenericOptionsParser;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
public class TimelineServicePerformance extends Configured implements Tool {
static final int NUM_MAPS_DEFAULT = 1;
static final int SIMPLE_ENTITY_WRITER = 1;
static final int JOB_HISTORY_FILE_REPLAY_MAPPER = 2;
static int mapperType = SIMPLE_ENTITY_WRITER;
static final int TIMELINE_SERVICE_VERSION_1 = 1;
static final int TIMELINE_SERVICE_VERSION_2 = 2;
static int timeline_service_version = TIMELINE_SERVICE_VERSION_1;
protected static int printUsage() {
System.err.println(
"Usage: [-m <maps>] number of mappers (default: " + NUM_MAPS_DEFAULT +
")\n" +
" [-v] timeline service version\n" +
" [-mtype <mapper type in integer>]\n" +
" 1. simple entity write mapper\n" +
" 2. jobhistory files replay mapper\n" +
" [-s <(KBs)test>] number of KB per put (mtype=1, default: " +
SimpleEntityWriterV1.KBS_SENT_DEFAULT + " KB)\n" +
" [-t] package sending iterations per mapper (mtype=1, default: " +
SimpleEntityWriterV1.TEST_TIMES_DEFAULT + ")\n" +
" [-d <path>] root path of job history files (mtype=2)\n" +
" [-r <replay mode>] (mtype=2)\n" +
" 1. write all entities for a job in one put (default)\n" +
" 2. write one entity at a time\n");
GenericOptionsParser.printGenericCommandUsage(System.err);
return -1;
}
/**
* Configure a job given argv.
*/
public static boolean parseArgs(String[] args, Job job) throws IOException {
// set the common defaults
Configuration conf = job.getConfiguration();
conf.setInt(MRJobConfig.NUM_MAPS, NUM_MAPS_DEFAULT);
for (int i = 0; i < args.length; i++) {
if (args.length == i + 1) {
System.out.println("ERROR: Required parameter missing from " + args[i]);
return printUsage() == 0;
}
try {
if ("-v".equals(args[i])) {
timeline_service_version = Integer.parseInt(args[++i]);
}
if ("-m".equals(args[i])) {
if (Integer.parseInt(args[++i]) > 0) {
job.getConfiguration()
.setInt(MRJobConfig.NUM_MAPS, Integer.parseInt(args[i]));
}
} else if ("-mtype".equals(args[i])) {
mapperType = Integer.parseInt(args[++i]);
} else if ("-s".equals(args[i])) {
if (Integer.parseInt(args[++i]) > 0) {
conf.setInt(SimpleEntityWriterV1.KBS_SENT, Integer.parseInt(args[i]));
}
} else if ("-t".equals(args[i])) {
if (Integer.parseInt(args[++i]) > 0) {
conf.setInt(SimpleEntityWriterV1.TEST_TIMES,
Integer.parseInt(args[i]));
}
} else if ("-d".equals(args[i])) {
conf.set(JobHistoryFileReplayHelper.PROCESSING_PATH, args[++i]);
} else if ("-r".equals(args[i])) {
conf.setInt(JobHistoryFileReplayHelper.REPLAY_MODE,
Integer.parseInt(args[++i]));
} else {
System.out.println("Unexpected argument: " + args[i]);
return printUsage() == 0;
}
} catch (NumberFormatException except) {
System.out.println("ERROR: Integer expected instead of " + args[i]);
return printUsage() == 0;
} catch (Exception e) {
throw (IOException)new IOException().initCause(e);
}
}
// handle mapper-specific settings
switch (timeline_service_version) {
case TIMELINE_SERVICE_VERSION_1:
default:
switch (mapperType) {
case JOB_HISTORY_FILE_REPLAY_MAPPER:
job.setMapperClass(JobHistoryFileReplayMapperV1.class);
String processingPath =
conf.get(JobHistoryFileReplayHelper.PROCESSING_PATH);
if (processingPath == null || processingPath.isEmpty()) {
System.out.println("processing path is missing while mtype = 2");
return printUsage() == 0;
}
break;
case SIMPLE_ENTITY_WRITER:
default:
job.setMapperClass(SimpleEntityWriterV1.class);
// use the current timestamp as the "run id" of the test: this will
// be used as simulating the cluster timestamp for apps
conf.setLong(SimpleEntityWriterV1.TIMELINE_SERVICE_PERFORMANCE_RUN_ID,
System.currentTimeMillis());
break;
}
}
return true;
}
/**
* TimelineServer Performance counters
*/
static enum PerfCounters {
TIMELINE_SERVICE_WRITE_TIME,
TIMELINE_SERVICE_WRITE_COUNTER,
TIMELINE_SERVICE_WRITE_FAILURES,
TIMELINE_SERVICE_WRITE_KBS,
}
public int run(String[] args) throws Exception {
Job job = Job.getInstance(getConf());
job.setJarByClass(TimelineServicePerformance.class);
job.setMapperClass(SimpleEntityWriterV1.class);
job.setInputFormatClass(SleepInputFormat.class);
job.setOutputFormatClass(NullOutputFormat.class);
job.setNumReduceTasks(0);
if (!parseArgs(args, job)) {
return -1;
}
Date startTime = new Date();
System.out.println("Job started: " + startTime);
int ret = job.waitForCompletion(true) ? 0 : 1;
org.apache.hadoop.mapreduce.Counters counters = job.getCounters();
long writetime =
counters.findCounter(PerfCounters.TIMELINE_SERVICE_WRITE_TIME).getValue();
long writecounts =
counters.findCounter(PerfCounters.TIMELINE_SERVICE_WRITE_COUNTER).getValue();
long writesize =
counters.findCounter(PerfCounters.TIMELINE_SERVICE_WRITE_KBS).getValue();
double transacrate = writecounts * 1000 / (double)writetime;
double iorate = writesize * 1000 / (double)writetime;
int numMaps =
Integer.parseInt(job.getConfiguration().get(MRJobConfig.NUM_MAPS));
System.out.println("TRANSACTION RATE (per mapper): " + transacrate +
" ops/s");
System.out.println("IO RATE (per mapper): " + iorate + " KB/s");
System.out.println("TRANSACTION RATE (total): " + transacrate*numMaps +
" ops/s");
System.out.println("IO RATE (total): " + iorate*numMaps + " KB/s");
return ret;
}
public static void main(String[] args) throws Exception {
int res =
ToolRunner.run(new Configuration(), new TimelineServicePerformance(),
args);
System.exit(res);
}
}

View File

@ -28,6 +28,7 @@ import org.apache.hadoop.mapred.TestMapRed;
import org.apache.hadoop.mapred.TestSequenceFileInputFormat; import org.apache.hadoop.mapred.TestSequenceFileInputFormat;
import org.apache.hadoop.mapred.TestTextInputFormat; import org.apache.hadoop.mapred.TestTextInputFormat;
import org.apache.hadoop.mapred.ThreadedMapBenchmark; import org.apache.hadoop.mapred.ThreadedMapBenchmark;
import org.apache.hadoop.mapreduce.TimelineServicePerformance;
import org.apache.hadoop.mapreduce.FailJob; import org.apache.hadoop.mapreduce.FailJob;
import org.apache.hadoop.mapreduce.LargeSorter; import org.apache.hadoop.mapreduce.LargeSorter;
import org.apache.hadoop.mapreduce.MiniHadoopClusterManager; import org.apache.hadoop.mapreduce.MiniHadoopClusterManager;
@ -90,6 +91,8 @@ public class MapredTestDriver {
pgd.addClass("fail", FailJob.class, "a job that always fails"); pgd.addClass("fail", FailJob.class, "a job that always fails");
pgd.addClass("sleep", SleepJob.class, pgd.addClass("sleep", SleepJob.class,
"A job that sleeps at each map and reduce task."); "A job that sleeps at each map and reduce task.");
pgd.addClass("timelineperformance", TimelineServicePerformance.class,
"A job that launches mappers to test timlineserver performance.");
pgd.addClass("nnbench", NNBench.class, pgd.addClass("nnbench", NNBench.class,
"A benchmark that stresses the namenode w/ MR."); "A benchmark that stresses the namenode w/ MR.");
pgd.addClass("nnbenchWithoutMR", NNBenchWithoutMR.class, pgd.addClass("nnbenchWithoutMR", NNBenchWithoutMR.class,

View File

@ -177,6 +177,9 @@ Release 2.8.0 - UNRELEASED
YARN-3964. Support NodeLabelsProvider at Resource Manager side. YARN-3964. Support NodeLabelsProvider at Resource Manager side.
(Dian Fu via devaraj) (Dian Fu via devaraj)
YARN-2556. Tool to measure the performance of the timeline server (Chang Li
via sjlee)
IMPROVEMENTS IMPROVEMENTS
YARN-644. Basic null check is not performed on passed in arguments before YARN-644. Basic null check is not performed on passed in arguments before