[
https://issues.apache.org/jira/browse/IGNITE-8890?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Puviarasu updated IGNITE-8890:
------------------------------
Description:
As Ignite-YARN is a long running application in YARN environment it should have
a mechanism to renew the delegation token.
In Ignite-YARN, when the ApplicationMaster is started, it acquires Delegation
tokens and stores in a ByteBuffer[Class: ApplicationMaster, Method: init()].
This ByteBuffer with token information is given to all the containers received
from ResourceManager [Class: ApplicationMaster, Method:
onContainersAllocated()].
Everything works fine till the life time of the delegation token.
Once the delegation token expires, the ApplicationMaster is not able to start
Ignite inside containers it receive and below exception occurs
*WARNING: Error launching container*
org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.security.token.SecretManager*$InvalidToken*)
:
at org.apache.hadoop.ipc.Client.call(Client.java:1504)
at org.apache.hadoop.ipc.Client.call(Client.java:1441)
at
org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:230)
at com.sun.proxy.$Proxy10.getFileInfo(Unknown Source)
at
org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.getFileInfo(ClientNamenodeProtocolTranslatorPB.java:771)
at sun.reflect.GeneratedMethodAccessor2.invoke(Unknown Source)
at
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at
org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:256)
at
org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:104)
at com.sun.proxy.$Proxy11.getFileInfo(Unknown Source)
at org.apache.hadoop.hdfs.DFSClient.getFileInfo(DFSClient.java:2123)
at
org.apache.hadoop.hdfs.DistributedFileSystem$20.doCall(DistributedFileSystem.java:1253)
at
org.apache.hadoop.hdfs.DistributedFileSystem$20.doCall(DistributedFileSystem.java:1249)
at
org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
at
org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1249)
at
org.apache.ignite.yarn.utils.IgniteYarnUtils.setupFile(IgniteYarnUtils.java:65)
at
org.apache.ignite.yarn.ApplicationMaster.onContainersAllocated(ApplicationMaster.java:131)
at
org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl$CallbackHandlerThread.run(AMRMClientAsyncImpl.java:292)
ApplicationMaster keeps on asking for more and more containers [Class:
ApplicationMaster, Method: run()] but not able to start Ignite inside any of
the containers due to the expired/missing delegation token. The failed
containers are not released when Exception occurs.
*This repeats until all the resources in the cluster are allocated to
Ignition. As a result of this Ignition uses all resources in the cluster and no
other jobs were able to run.*
Kindly help in resolving the issue.
Thanks in Advance!!!
was:
As Ignite-YARN is a long running application in YARN environment it should have
a mechanism to renew the delegation token.
In Ignite-YARN, when the ApplicationMaster is started, it acquires Delegation
tokens and stores in a ByteBuffer[Class: ApplicationMaster, Method: init()].
This ByteBuffer with token information is given to all the containers received
from ResourceManager [Class: ApplicationMaster, Method:
onContainersAllocated()].
Everything works fine till the life time of the delegation token.
Once the delegation token expires, the ApplicationMaster is not able to start
Ignite inside containers it receive and below exception occurs
WARNING: Error launching container
org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.security.token.SecretManager$InvalidToken):
at org.apache.hadoop.ipc.Client.call(Client.java:1504)
at org.apache.hadoop.ipc.Client.call(Client.java:1441)
at
org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:230)
at com.sun.proxy.$Proxy10.getFileInfo(Unknown Source)
at
org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.getFileInfo(ClientNamenodeProtocolTranslatorPB.java:771)
at sun.reflect.GeneratedMethodAccessor2.invoke(Unknown Source)
at
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at
org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:256)
at
org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:104)
at com.sun.proxy.$Proxy11.getFileInfo(Unknown Source)
at org.apache.hadoop.hdfs.DFSClient.getFileInfo(DFSClient.java:2123)
at
org.apache.hadoop.hdfs.DistributedFileSystem$20.doCall(DistributedFileSystem.java:1253)
at
org.apache.hadoop.hdfs.DistributedFileSystem$20.doCall(DistributedFileSystem.java:1249)
at
org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
at
org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1249)
at
org.apache.ignite.yarn.utils.IgniteYarnUtils.setupFile(IgniteYarnUtils.java:65)
at
org.apache.ignite.yarn.ApplicationMaster.onContainersAllocated(ApplicationMaster.java:131)
at
org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl$CallbackHandlerThread.run(AMRMClientAsyncImpl.java:292)
ApplicationMaster keeps on asking for more and more containers [Class:
ApplicationMaster, Method: run()] but not able to start Ignite inside any of
the containers due to the expired/missing delegation token.
This repeats until all the resources in the cluster are allocated to Ignition.
Kindly help in resolving the issue.
Thanks in Advance!!!
> Ignite YARN Kerberos - Delegation Token renewal
> -----------------------------------------------
>
> Key: IGNITE-8890
> URL: https://issues.apache.org/jira/browse/IGNITE-8890
> Project: Ignite
> Issue Type: Bug
> Components: yarn
> Affects Versions: 2.3
> Environment: Kerberos cluster
> Ignite Version : 2.3.0
> Module : Ignite-YARN
> Class : ApplicationMaster
>
> Reporter: Puviarasu
> Priority: Blocker
>
> As Ignite-YARN is a long running application in YARN environment it should
> have a mechanism to renew the delegation token.
> In Ignite-YARN, when the ApplicationMaster is started, it acquires Delegation
> tokens and stores in a ByteBuffer[Class: ApplicationMaster, Method: init()].
> This ByteBuffer with token information is given to all the containers
> received from ResourceManager [Class: ApplicationMaster, Method:
> onContainersAllocated()].
> Everything works fine till the life time of the delegation token.
> Once the delegation token expires, the ApplicationMaster is not able to start
> Ignite inside containers it receive and below exception occurs
> *WARNING: Error launching container*
>
> org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.security.token.SecretManager*$InvalidToken*)
> :
> at org.apache.hadoop.ipc.Client.call(Client.java:1504)
> at org.apache.hadoop.ipc.Client.call(Client.java:1441)
> at
> org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:230)
> at com.sun.proxy.$Proxy10.getFileInfo(Unknown Source)
> at
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.getFileInfo(ClientNamenodeProtocolTranslatorPB.java:771)
> at sun.reflect.GeneratedMethodAccessor2.invoke(Unknown Source)
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:498)
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:256)
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:104)
> at com.sun.proxy.$Proxy11.getFileInfo(Unknown Source)
> at org.apache.hadoop.hdfs.DFSClient.getFileInfo(DFSClient.java:2123)
> at
> org.apache.hadoop.hdfs.DistributedFileSystem$20.doCall(DistributedFileSystem.java:1253)
> at
> org.apache.hadoop.hdfs.DistributedFileSystem$20.doCall(DistributedFileSystem.java:1249)
> at
> org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
> at
> org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1249)
> at
> org.apache.ignite.yarn.utils.IgniteYarnUtils.setupFile(IgniteYarnUtils.java:65)
> at
> org.apache.ignite.yarn.ApplicationMaster.onContainersAllocated(ApplicationMaster.java:131)
> at
> org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl$CallbackHandlerThread.run(AMRMClientAsyncImpl.java:292)
> ApplicationMaster keeps on asking for more and more containers [Class:
> ApplicationMaster, Method: run()] but not able to start Ignite inside any of
> the containers due to the expired/missing delegation token. The failed
> containers are not released when Exception occurs.
> *This repeats until all the resources in the cluster are allocated to
> Ignition. As a result of this Ignition uses all resources in the cluster and
> no other jobs were able to run.*
> Kindly help in resolving the issue.
> Thanks in Advance!!!
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
