[jira] [Updated] (HADOOP-11000) HAServiceProtocol's health state is incorrectly transitioned to SERVICE_NOT_RESPONDING

["Ming Ma (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HADOOP-11000?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ming Ma updated HADOOP-11000:
-----------------------------

    Attachment: HADOOP-11000.patch

Here is the draft patch.

1. Catch HealthCheckFailedException properly.
2. Update test code for HA scenario to cover protocol buffer RPC scenario.

> HAServiceProtocol's health state is incorrectly transitioned to SERVICE_NOT_RESPONDING
> --------------------------------------------------------------------------------------
>
>                 Key: HADOOP-11000
>                 URL: https://issues.apache.org/jira/browse/HADOOP-11000
>             Project: Hadoop Common
>          Issue Type: Bug
>            Reporter: Ming Ma
>         Attachments: HADOOP-11000.patch
>
>
> When HAServiceProtocol.monitorHealth throws a HealthCheckFailedException, the actual exception from protocol buffer RPC is a RemoteException that wraps the real exception. Thus the state is incorrectly transitioned to SERVICE_NOT_RESPONDING
> {noformat}
> HealthMonitor.java
> doHealthChecks
>       try {
>         status = proxy.getServiceStatus();
>         proxy.monitorHealth();
>         healthy = true;
>       } catch (HealthCheckFailedException e) {
>         .....
>         enterState(State.SERVICE_UNHEALTHY);
>       } catch (Throwable t) {
>         .....
>         enterState(State.SERVICE_NOT_RESPONDING);
>         .....
>       }
> {noformat}



--
This message was sent by Atlassian JIRA
(v6.2#6252)