big-data/hadoop
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

MAPREDUCE-6335. Created MR job based performance test driver for the timeline service v2. Contributed by Sangjin Lee.

Browse Source 
(cherry picked from commit b689f5d43d3f5434a30fe52f1a7e12e1fc5c71f4)
This commit is contained in:
Zhijie Shen 2015-04-28 19:46:01 -07:00 committed by Sangjin Lee
parent b50a6d78f5
commit 8c7b6dd2c7
2 changed files with 302 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

298
hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/TimelineServicePerformanceV2.java Normal file
View File

@ -0,0 +1,298 @@
/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.mapred;
import java.io.IOException;
import java.util.Date;
import java.util.Random;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.MRJobConfig;
import org.apache.hadoop.mapreduce.SleepJob.SleepInputFormat;
import org.apache.hadoop.mapreduce.TaskAttemptID;
import org.apache.hadoop.mapreduce.lib.output.NullOutputFormat;
import org.apache.hadoop.security.UserGroupInformation;
import org.apache.hadoop.util.GenericOptionsParser;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.apache.hadoop.yarn.api.records.ApplicationId;
import org.apache.hadoop.yarn.api.records.timelineservice.TimelineEntities;
import org.apache.hadoop.yarn.api.records.timelineservice.TimelineEntity;
import org.apache.hadoop.yarn.api.records.timelineservice.TimelineEvent;
import org.apache.hadoop.yarn.api.records.timelineservice.TimelineMetric;
import org.apache.hadoop.yarn.conf.YarnConfiguration;
import org.apache.hadoop.yarn.server.timelineservice.collector.AppLevelTimelineCollector;
import org.apache.hadoop.yarn.server.timelineservice.collector.TimelineCollectorContext;
public class TimelineServicePerformanceV2 extends Configured implements Tool {
private static final Log LOG =
LogFactory.getLog(TimelineServicePerformanceV2.class);
static final int NUM_MAPS_DEFAULT = 1;
static final int SIMPLE_ENTITY_WRITER = 1;
// 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";
static int mapperType = SIMPLE_ENTITY_WRITER;
protected static int printUsage() {
// TODO is there a way to handle mapper-specific options more gracefully?
System.err.println(
"Usage: [-m <maps>] number of mappers (default: " + NUM_MAPS_DEFAULT +
")\n" +
" [-mtype <mapper type in integer>] \n" +
" 1. simple entity write mapper\n" +
" [-s <(KBs)test>] number of KB per put (default: " +
KBS_SENT_DEFAULT + " KB)\n" +
" [-t] package sending iterations per mapper (default: " +
TEST_TIMES_DEFAULT + ")\n");
GenericOptionsParser.printGenericCommandUsage(System.err);
return -1;
}
/**
* Configure a job given argv.
*/
public static boolean parseArgs(String[] args, Job job) throws IOException {
// set the defaults
Configuration conf = job.getConfiguration();
conf.setInt(MRJobConfig.NUM_MAPS, NUM_MAPS_DEFAULT);
conf.setInt(KBS_SENT, KBS_SENT_DEFAULT);
conf.setInt(TEST_TIMES, TEST_TIMES_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 ("-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]);
switch (mapperType) {
case SIMPLE_ENTITY_WRITER:
job.setMapperClass(SimpleEntityWriter.class);
break;
default:
job.setMapperClass(SimpleEntityWriter.class);
}
} else if ("-s".equals(args[i])) {
if (Integer.parseInt(args[++i]) > 0) {
conf.setInt(KBS_SENT, (Integer.parseInt(args[i])));
}
} else if ("-t".equals(args[i])) {
if (Integer.parseInt(args[++i]) > 0) {
conf.setInt(TEST_TIMES, (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);
}
}
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(TimelineServicePerformanceV2.class);
job.setMapperClass(SimpleEntityWriter.class);
job.setInputFormatClass(SleepInputFormat.class);
job.setOutputFormatClass(NullOutputFormat.class);
job.setNumReduceTasks(0);
if (!parseArgs(args, job)) {
return -1;
}
// for mtype = 1
// use the current timestamp as the "run id" of the test: this will be used
// as simulating the cluster timestamp for apps
Configuration conf = job.getConfiguration();
conf.setLong(TIMELINE_SERVICE_PERFORMANCE_RUN_ID,
System.currentTimeMillis());
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(conf.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 TimelineServicePerformanceV2(),
args);
System.exit(res);
}
/**
* To ensure that the compression really gets exercised, generate a
* random alphanumeric fixed length payload
*/
static final char[] alphaNums = 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', ' ' };
/**
* Adds simple entities with random string payload, events, metrics, and
* configuration.
*/
public static class SimpleEntityWriter
extends org.apache.hadoop.mapreduce.Mapper<IntWritable,IntWritable,Writable,Writable> {
public void map(IntWritable key, IntWritable val, Context context)
throws IOException {
Configuration conf = context.getConfiguration();
// simulate the app id with the task id
int taskId = context.getTaskAttemptID().getTaskID().getId();
long timestamp = conf.getLong(TIMELINE_SERVICE_PERFORMANCE_RUN_ID, 0);
ApplicationId appId = ApplicationId.newInstance(timestamp, taskId);
// create the app level timeline collector
Configuration tlConf = new YarnConfiguration();
AppLevelTimelineCollector collector =
new AppLevelTimelineCollector(appId);
collector.init(tlConf);
collector.start();
try {
// set the context
// flow id: job name, flow run id: timestamp, user id
TimelineCollectorContext tlContext =
collector.getTimelineEntityContext();
tlContext.setFlowName(context.getJobName());
tlContext.setFlowRunId(timestamp);
tlContext.setUserId(context.getUser());
final int kbs = Integer.parseInt(conf.get(KBS_SENT));
long totalTime = 0;
final int testtimes = Integer.parseInt(conf.get(TEST_TIMES));
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(alphaNums.length);
payLoad[xx] = alphaNums[alphaNumIdx];
}
String entId = taskAttemptId + "_" + Integer.toString(i);
final TimelineEntity entity = new TimelineEntity();
entity.setId(entId);
entity.setType("FOO_ATTEMPT");
entity.addInfo("PERF_TEST", payLoad);
// add an event
TimelineEvent event = new TimelineEvent();
event.setTimestamp(System.currentTimeMillis());
event.addInfo("foo_event", "test");
entity.addEvent(event);
// add a metric
TimelineMetric metric = new TimelineMetric();
metric.setId("foo_metric");
metric.setSingleData(123456789L);
entity.addMetric(metric);
// add a config
entity.addConfig("foo", "bar");
TimelineEntities entities = new TimelineEntities();
entities.addEntity(entity);
// use the current user for this purpose
UserGroupInformation ugi = UserGroupInformation.getCurrentUser();
long startWrite = System.nanoTime();
try {
collector.putEntities(entities, ugi);
} catch (Exception e) {
context.getCounter(PerfCounters.TIMELINE_SERVICE_WRITE_FAILURES).
increment(1);
e.printStackTrace();
}
long endWrite = System.nanoTime();
totalTime += (endWrite-startWrite)/1000000L;
}
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);
} finally {
// clean up
collector.close();
}
}
}
}

5
hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/test/MapredTestDriver.java
View File

@ -29,6 +29,7 @@
import org.apache.hadoop.mapred.TestTextInputFormat;
import org.apache.hadoop.mapred.ThreadedMapBenchmark;
import org.apache.hadoop.mapreduce.TimelineServicePerformance;
import org.apache.hadoop.mapred.TimelineServicePerformanceV2;
import org.apache.hadoop.mapreduce.FailJob;
import org.apache.hadoop.mapreduce.LargeSorter;
import org.apache.hadoop.mapreduce.MiniHadoopClusterManager;
@ -93,7 +94,9 @@ public MapredTestDriver(ProgramDriver pgd) {
"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("timelineperformance", TimelineServicePerformanceV2.class,
"A job that launch mappers to test timline service v.2 performance.");
pgd.addClass("nnbench", NNBench.class,
"A benchmark that stresses the namenode w/ MR.");
pgd.addClass("nnbenchWithoutMR", NNBenchWithoutMR.class,
"A benchmark that stresses the namenode w/o MR.");