[
https://issues.apache.org/jira/browse/HADOOP-4659?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hairong Kuang updated HADOOP-4659:
----------------------------------
Attachment: rpcConn.patch
This patch checks the cause of the failure when setting up a RPC client tries
to connect to a RPC server. It retries if it is caused by an unavailable or
busy server.
It adds a new static method waitForProxy with a timeout mainly for the purpose
of testing. A unit test is added to TestRPC to makes sure that client retries.
A manual test is also conducted that starting a DataNode without starting
NameNode causes DataNode to retry.
Steve, could you please review and test the patch in your setup? I appreciate
any of your feedback.
> Root cause of connection failure is being lost to code that uses it for
> delaying startup
> ----------------------------------------------------------------------------------------
>
> Key: HADOOP-4659
> URL: https://issues.apache.org/jira/browse/HADOOP-4659
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.3
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Priority: Blocker
> Fix For: 0.18.3
>
> Attachments: connectRetry.patch, hadoop-4659.patch, rpcConn.patch
>
>
> ipc.Client the root cause of a connection failure is being lost as the
> exception is wrapped, hence the outside code, the one that looks for that
> root cause, isn't working as expected. The results is you can't bring up a
> task tracker before job tracker, and probably the same for a datanode before
> a namenode. The change that triggered this is not yet located, I had thought
> it was HADOOP-3844 but I no longer believe this is the case.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
