You are viewing a plain text version of this content. The canonical link for it is here.
Posted to yarn-issues@hadoop.apache.org by "Zhijie Shen (JIRA)" <ji...@apache.org> on 2014/11/19 20:48:33 UTC
[jira] [Commented] (YARN-2879) Compatibility validation between
YARN 2.2/2.4 and 2.6
[ https://issues.apache.org/jira/browse/YARN-2879?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14218388#comment-14218388 ]
Zhijie Shen commented on YARN-2879:
-----------------------------------
a. In the following scenarios:
1. Either insecure or secure;
2. MR 2.2 with either old or new shuffle handler on NM;
3. Submitting via new client.
We will see the following console exception:
{code}
14/11/17 23:47:45 INFO mapreduce.JobSubmitter: Cleaning up the staging area /user/zjshen/.staging/zjshen/.staging/job_1416270549965_0014
java.lang.NoSuchMethodError: org.apache.hadoop.http.HttpConfig.getSchemePrefix()Ljava/lang/String;
at org.apache.hadoop.mapred.ClientServiceDelegate.getJobStatus(ClientServiceDelegate.java:428)
at org.apache.hadoop.mapred.YARNRunner.submitJob(YARNRunner.java:302)
at org.apache.hadoop.mapreduce.JobSubmitter.submitJobInternal(JobSubmitter.java:430)
at org.apache.hadoop.mapreduce.Job$10.run(Job.java:1268)
at org.apache.hadoop.mapreduce.Job$10.run(Job.java:1265)
at java.security.AccessController.doPrivileged(Native Method)
at javax.security.auth.Subject.doAs(Subject.java:415)
at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1628)
at org.apache.hadoop.mapreduce.Job.submit(Job.java:1265)
at org.apache.hadoop.mapreduce.Job.waitForCompletion(Job.java:1286)
at org.apache.hadoop.examples.QuasiMonteCarlo.estimatePi(QuasiMonteCarlo.java:306)
at org.apache.hadoop.examples.QuasiMonteCarlo.run(QuasiMonteCarlo.java:354)
at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:70)
at org.apache.hadoop.examples.QuasiMonteCarlo.main(QuasiMonteCarlo.java:363)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at org.apache.hadoop.util.ProgramDriver$ProgramDescription.invoke(ProgramDriver.java:71)
at org.apache.hadoop.util.ProgramDriver.run(ProgramDriver.java:144)
at org.apache.hadoop.examples.ExampleDriver.main(ExampleDriver.java:74)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at org.apache.hadoop.util.RunJar.run(RunJar.java:221)
at org.apache.hadoop.util.RunJar.main(RunJar.java:136)
{code}
b. In the following scenarios:
1. Either insecure or secure;
2. MR 2.2 with old on NM;
3. Submitting via old client.
We will see the following exception in the AM Log:
{code}
2014-11-17 15:09:06,157 INFO [main] org.apache.hadoop.mapreduce.v2.app.MRAppMaster: Created MRAppMaster for application appattempt_1416264695865_0007_000001
2014-11-17 15:09:06,436 FATAL [main] org.apache.hadoop.mapreduce.v2.app.MRAppMaster: Error starting MRAppMaster
java.lang.NoSuchMethodError: org.apache.hadoop.http.HttpConfig.setPolicy(Lorg/apache/hadoop/http/HttpConfig$Policy;)V
at org.apache.hadoop.mapreduce.v2.app.MRAppMaster.main(MRAppMaster.java:1364)
2014-11-17 15:09:06,439 INFO [Thread-1] org.apache.hadoop.mapreduce.v2.app.MRAppMaster: MRAppMaster received a signal. Signaling RMCommunicator and JobHistoryEventHandler.
{code}
The two exceptions are actually the same problem, but using the old client prevents it happening during app submission. Will file a separate Jira for it.
> Compatibility validation between YARN 2.2/2.4 and 2.6
> -----------------------------------------------------
>
> Key: YARN-2879
> URL: https://issues.apache.org/jira/browse/YARN-2879
> Project: Hadoop YARN
> Issue Type: Bug
> Reporter: Zhijie Shen
> Assignee: Zhijie Shen
>
> Recently, I did some simple backward compatibility experiments. Bascially, I've taken the following 2 steps:
> 1. Deploy the application (MR and DistributedShell) that is compiled against *old* YARN API (2.2/2.4) on *new* YARN cluster (2.6). The application is submitted via *new* Hadoop (2.6) client.
> 2. Deploy the application (MR and DistributedShell) that is compiled against *old* YARN API (2.2/2.4) on *new* YARN cluster (2.6). The application is submitted via *old* Hadoop (2.2/2.4) client that comes with the app.
> I've tried these 2 steps on both insecure and secure cluster. Here's a short summary:
> || || || DS 2.2 || DS 2.4 || MR 2.2 + Shuffle 2.2 || MR 2.2 + Shuffle 2.6 || MR 2.4 + Shuffle 2.4 || MR 2.4 + Shuffle 2.6 ||
> | Insecure | New Client | OK | OK | Client Incompatible | Client Incompatible | OK | OK |
> | Insecure | Old Client | OK | OK | AM Incompatible | Client Incompatible | OK | OK |
> | secure | New Client | OK | OK | Client Incompatible | Client Incompatible | OK | OK |
> | secure | Old Client | OK | OK | AM Incompatible | Client Incompatible | OK | OK |
> Note that I've tried to run NM with both old and new shuffle handler version.
> In general, the compatibility looks good overall. There're a few issues that are related to MR, but they seem to be not the YARN issue. I'll post the individual problem in the follow-up comments.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)