[
https://issues.apache.org/jira/browse/TEZ-3716?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Siddharth Seth updated TEZ-3716:
--------------------------------
Attachment: TEZ-3716.01.patch
Switched from changing the dagschedulers to making the change in
TaskAttemptImpl itself, which is where the final priority is selected.
Little larger than expected since this introduces a Config in Vertex to avoid
reading the config from hadoop configuration for each attempt.
> Allow attempt retries to be treated the same as the first attempt
> -----------------------------------------------------------------
>
> Key: TEZ-3716
> URL: https://issues.apache.org/jira/browse/TEZ-3716
> Project: Apache Tez
> Issue Type: Task
> Reporter: Siddharth Seth
> Assignee: Siddharth Seth
> Attachments: TEZ-3716.01.patch
>
>
> Essentially - don't change attempt priority due to a failure.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
