You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@helix.apache.org by "Kanak Biscuitwala (JIRA)" <ji...@apache.org> on 2014/05/27 23:03:02 UTC

[jira] [Resolved] (HELIX-438) Improve task framework retry logic

     [ https://issues.apache.org/jira/browse/HELIX-438?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Kanak Biscuitwala resolved HELIX-438.
-------------------------------------

    Resolution: Fixed
      Assignee: Kanak Biscuitwala

The general way that this works is that there are up to M total retries that Helix has to play with. The default retry strategy will partition those M into M/N blocks, where N is the maximum number of times Helix is allowed to force a reassignment. So every N/M tries, there will be a reassignment of failing tasks.

This logic is encapsulated within a new interface called RetryPolicy, and future changes can be made to make this retry policy pluggable.

Review: https://reviews.apache.org/r/21879/

> Improve task framework retry logic
> ----------------------------------
>
>                 Key: HELIX-438
>                 URL: https://issues.apache.org/jira/browse/HELIX-438
>             Project: Apache Helix
>          Issue Type: Sub-task
>            Reporter: Kanak Biscuitwala
>            Assignee: Kanak Biscuitwala
>
> Support up to M local retries, N retries elsewhere (per task, configured per job).



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