[
https://issues.apache.org/jira/browse/TEZ-853?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14127830#comment-14127830
]
Jeff Zhang commented on TEZ-853:
--------------------------------
[~hitesh]
bq. Why is job1 being compared against job2? Based on the kind of failing dag
being run, should the test be able to verify how many failures and/or kills are
expected? Also, totalSuccess should be calculated based on total num tasks
across vertices - correct? Is there a counter for succeeded tasks - if yes,
that should be verified too.
Because of the limitation of the VM (FailOnAttemptVertexManagerPlugin), most of
time there's no task running when VM kill itself . So actually there's no task
failed/killed. I have designed a new VM which could kill AM when task is
running in [Tez-1559|https://issues.apache.org/jira/browse/TEZ-1559]. Maybe I
could include it in this jira.
bq. Is there a counter for succeeded tasks - if yes, that should be verified
too.
There's no counter for succeeded tasks. That's why calculate it based on the
total_launched, total_killed, and total_failed.
> Support counters recovery
> -------------------------
>
> Key: TEZ-853
> URL: https://issues.apache.org/jira/browse/TEZ-853
> Project: Apache Tez
> Issue Type: Sub-task
> Reporter: Hitesh Shah
> Assignee: Jeff Zhang
> Attachments: Tez-853-2.patch, Tez-853.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
