hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gary Helmling (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-3206) FairScheduler breaks writing of job tokens to MR system dir when using kerberos auth
Date Wed, 19 Oct 2011 18:43:12 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-3206?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13130865#comment-13130865
] 

Gary Helmling commented on MAPREDUCE-3206:
------------------------------------------

I'm not sure pushing down the remote user UGI.doAs() is really possible.  As I understand
it, the current problem is the UGI.doAs() wrapping the entire RPC call invocation up in org.apache.hadoop.ipc.Server$Handler.run():
{code}
CurCall.set(call);
try {
  // Make the call as the user via Subject.doAs, thus associating
  // the call with the Subject
  if (call.connection.user == null) {
    value = call(call.connection.protocol, call.param, 
                 call.timestamp);
  } else {
    value = 
        call.connection.user.doAs
            (new PrivilegedExceptionAction<Writable>() {
               @Override
               public Writable run() throws Exception {
                 // make the call
                 return call(call.connection.protocol, 
                             call.param, call.timestamp);

               }
             }
            );
  }
{code}

Prior to MAPREDUCE-2981, as I understand it, JobTracker.initJob() was being invoked in a separate
thread by the scheduler, not under the RPC handler thread.  So it was picking up the JobTracker
login user since it wasn't running under a UGI.doAs() block.

I haven't checked the FairScheduler code in 0.23 or trunk to see how it's handled there. 
Maybe that would provide another alternative.  Or changing FairScheduler$JobInitializer to
run in a separate thread and pull jobs to initialize from a queue might work as well.  Maybe
that's more along the lines of what you were thinking?  I just tried the least intrusive approach.
                
> FairScheduler breaks writing of job tokens to MR system dir when using kerberos auth
> ------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-3206
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-3206
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: contrib/fair-share
>    Affects Versions: 0.20.205.0
>            Reporter: Gary Helmling
>            Priority: Critical
>         Attachments: MAPREDUCE-3206_0.20.205.patch
>
>
> Using the FairScheduler with kerberos authentication does not appear to work with 0.20.205.0.
 When submitting a job, execution fails when storing the delegation tokens for the job:
> {noformat}
> 11/10/18 21:48:53 INFO hdfs.DFSClient: Created HDFS_DELEGATION_TOKEN token 6 for hdfs
on xxxx:8020
> 11/10/18 21:48:53 INFO security.TokenCache: Got dt for hdfs://xxxx/user/hdfs/.staging/job_201110182148_0001;uri=xxxx:8020;t.service=xxxx:8020
> 11/10/18 21:48:53 INFO mapred.FileInputFormat: Total input paths to process : 1
> 11/10/18 21:48:53 INFO mapred.JobClient: Running job: job_201110182148_0001
> 11/10/18 21:48:54 INFO mapred.JobClient:  map 0% reduce 0%
> 11/10/18 21:48:54 INFO mapred.JobClient: Job complete: job_201110182148_0001
> 11/10/18 21:48:54 INFO mapred.JobClient: Counters: 0
> 11/10/18 21:48:54 INFO mapred.JobClient: Job Failed: Job initialization failed:
> java.io.IOException: Call to xxxx/xxxx:8020 failed on local exception: java.io.IOException:
javax.security.sasl.SaslException: GSS initiate failed [Caused by GSSException: No valid credentials
provided (Mechanism level: Attempt to obtain new INITIATE credentials failed! (null))]
> 	at org.apache.hadoop.ipc.Client.wrapException(Client.java:1103)
> 	at org.apache.hadoop.ipc.Client.call(Client.java:1071)
> 	at org.apache.hadoop.ipc.RPC$Invoker.invoke(RPC.java:225)
> 	at $Proxy7.getProtocolVersion(Unknown Source)
> 	at org.apache.hadoop.ipc.RPC.getProxy(RPC.java:396)
> 	at org.apache.hadoop.ipc.RPC.getProxy(RPC.java:379)
> 	at org.apache.hadoop.hdfs.DFSClient.createRPCNamenode(DFSClient.java:118)
> 	at org.apache.hadoop.hdfs.DFSClient.<init>(DFSClient.java:222)
> 	at org.apache.hadoop.hdfs.DFSClient.<init>(DFSClient.java:187)
> 	at org.apache.hadoop.hdfs.DistributedFileSystem.initialize(DistributedFileSystem.java:89)
> 	at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:1328)
> 	at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:65)
> 	at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:1346)
> 	at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:244)
> 	at org.apache.hadoop.fs.Path.getFileSystem(Path.java:187)
> 	at org.apache.hadoop.security.Credentials.writeTokenStorageFile(Credentials.java:175)
> 	at org.apache.hadoop.mapred.JobInProgress.generateAndStoreTokens(JobInProgress.java:3528)
> 	at org.apache.hadoop.mapred.JobInProgress.initTasks(JobInProgress.java:696)
> 	at org.apache.hadoop.mapred.JobTracker.initJob(JobTracker.java:4207)
> 	at org.apache.hadoop.mapred.FairScheduler$JobInitializer$InitJob.run(FairScheduler.java:291)
> 	at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
> 	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
> 	at java.lang.Thread.run(Thread.java:662)
> Caused by: java.io.IOException: javax.security.sasl.SaslException: GSS initiate failed
[Caused by GSSException: No valid credentials provided (Mechanism level: Attempt to obtain
new INITIATE credentials failed! (null))]
> 	at org.apache.hadoop.ipc.Client$Connection$1.run(Client.java:539)
> 	at java.security.AccessController.doPrivileged(Native Method)
> 	at javax.security.auth.Subject.doAs(Subject.java:396)
> 	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1059)
> 	at org.apache.hadoop.ipc.Client$Connection.handleSaslConnectionFailure(Client.java:484)
> 	at org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:586)
> 	at org.apache.hadoop.ipc.Client$Connection.access$2000(Client.java:184)
> 	at org.apache.hadoop.ipc.Client.getConnection(Client.java:1202)
> 	at org.apache.hadoop.ipc.Client.call(Client.java:1046)
> 	... 21 more
> Caused by: javax.security.sasl.SaslException: GSS initiate failed [Caused by GSSException:
No valid credentials provided (Mechanism level: Attempt to obtain new INITIATE credentials
failed! (null))]
> 	at com.sun.security.sasl.gsskerb.GssKrb5Client.evaluateChallenge(GssKrb5Client.java:194)
> 	at org.apache.hadoop.security.SaslRpcClient.saslConnect(SaslRpcClient.java:134)
> 	at org.apache.hadoop.ipc.Client$Connection.setupSaslConnection(Client.java:381)
> 	at org.apache.hadoop.ipc.Client$Connection.access$1100(Client.java:184)
> 	at org.apache.hadoop.ipc.Client$Connection$2.run(Client.java:579)
> 	at org.apache.hadoop.ipc.Client$Connection$2.run(Client.java:576)
> 	at java.security.AccessController.doPrivileged(Native Method)
> 	at javax.security.auth.Subject.doAs(Subject.java:396)
> 	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1059)
> 	at org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:575)
> 	... 24 more
> Caused by: GSSException: No valid credentials provided (Mechanism level: Attempt to obtain
new INITIATE credentials failed! (null))
> 	at sun.security.jgss.krb5.Krb5InitCredential.getTgt(Krb5InitCredential.java:333)
> 	at sun.security.jgss.krb5.Krb5InitCredential.getInstance(Krb5InitCredential.java:128)
> 	at sun.security.jgss.krb5.Krb5MechFactory.getCredentialElement(Krb5MechFactory.java:106)
> 	at sun.security.jgss.krb5.Krb5MechFactory.getMechanismContext(Krb5MechFactory.java:172)
> 	at sun.security.jgss.GSSManagerImpl.getMechanismContext(GSSManagerImpl.java:209)
> 	at sun.security.jgss.GSSContextImpl.initSecContext(GSSContextImpl.java:195)
> 	at sun.security.jgss.GSSContextImpl.initSecContext(GSSContextImpl.java:162)
> 	at com.sun.security.sasl.gsskerb.GssKrb5Client.evaluateChallenge(GssKrb5Client.java:175)
> 	... 33 more
> Caused by: javax.security.auth.login.LoginException: No LoginModules configured for 
> 	at javax.security.auth.login.LoginContext.init(LoginContext.java:256)
> 	at javax.security.auth.login.LoginContext.<init>(LoginContext.java:499)
> 	at sun.security.jgss.GSSUtil.login(GSSUtil.java:244)
> 	at sun.security.jgss.krb5.Krb5Util.getTicket(Krb5Util.java:136)
> 	at sun.security.jgss.krb5.Krb5InitCredential$1.run(Krb5InitCredential.java:328)
> 	at java.security.AccessController.doPrivileged(Native Method)
> 	at sun.security.jgss.krb5.Krb5InitCredential.getTgt(Krb5InitCredential.java:325)
> 	... 40 more
> {noformat}
> The problem seems to have been introduced by the backported changes in MAPREDUCE-2981,
which shifted the execution of JobTracker.initJob(), and hence JobInProgress.generateAndStoreTokens(),
to underneath the call path for the RPC invocation.  As a result, the DFS write in TokenStorage.writeTokenStorageFile()
in done under a UGI.doAs() block as the RPC client remote user, without a TGT for negotiating
the connection.
> Does this analysis seem right?  Previously it seems that JobTracker.initJob() was only
called in a separate thread so it was picking up the credentials obtained for the configured
JobTracker kerberos principal.  The same job runs successfully in a build with MAPREDUCE-2981
reverted.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message