You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-issues@hadoop.apache.org by "xuzq (JIRA)" <ji...@apache.org> on 2019/06/17 09:45:00 UTC
[jira] [Updated] (HDFS-14543) RBF: RouterRpcClient shouldn't retry
when the remoteException is RetryException
[ https://issues.apache.org/jira/browse/HDFS-14543?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
xuzq updated HDFS-14543:
------------------------
Description:
Currently RouterRpcClient will retry when the remote exception is RetryException.
And it will use one handler to retry 3 times, if the remote exception is always RetryException.
Although there is no sleep between retries, but it may costs some other resources.
So, RouterRpcClient show return the RetryException to client if the name node returns RetryException, and do sleep at client.
After sleep some times, client can request the same Router.
was:
Currently RouterRpcClient will retry when the remote exception is RetryException.
And it will use one handler to retry 15 times, if the remote exception is always RetryException.
Although there is no sleep between retries, but it may costs some other resources.
So, RouterRpcClient show return the RetryException to client if the name node returns RetryException, and do sleep at client.
After sleep some times, client can request the same Router.
> RBF: RouterRpcClient shouldn't retry when the remoteException is RetryException
> -------------------------------------------------------------------------------
>
> Key: HDFS-14543
> URL: https://issues.apache.org/jira/browse/HDFS-14543
> Project: Hadoop HDFS
> Issue Type: Task
> Components: rbf
> Reporter: xuzq
> Assignee: xuzq
> Priority: Major
>
> Currently RouterRpcClient will retry when the remote exception is RetryException.
> And it will use one handler to retry 3 times, if the remote exception is always RetryException.
> Although there is no sleep between retries, but it may costs some other resources.
> So, RouterRpcClient show return the RetryException to client if the name node returns RetryException, and do sleep at client.
> After sleep some times, client can request the same Router.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org