You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@helix.apache.org by "Zhen Zhang (JIRA)" <ji...@apache.org> on 2014/10/01 23:41:34 UTC

[jira] [Commented] (HELIX-518) Add integration tests to ensure helix tasks work as expected during master failover

    [ https://issues.apache.org/jira/browse/HELIX-518?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14155599#comment-14155599 ] 

Zhen Zhang commented on HELIX-518:
----------------------------------

[~kanak] can we call TaskRunner#cancel or Task#cancel directly? I see it can be called either during state transition or timeout. If we call it directly, then the task will be left in RUNNING state?

> Add integration tests to ensure helix tasks work as expected during master failover
> -----------------------------------------------------------------------------------
>
>                 Key: HELIX-518
>                 URL: https://issues.apache.org/jira/browse/HELIX-518
>             Project: Apache Helix
>          Issue Type: Bug
>            Reporter: Zhen Zhang
>            Assignee: Zhen Zhang
>   Original Estimate: 72h
>  Remaining Estimate: 72h
>
> Integration test to verify that a helix task running on master gets rescheduled to new master during mastership handoff
> Integration test to verify that a helix task running on master gets rescheduled to new master when the current master just dies



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)