[
https://issues.apache.org/jira/browse/HAWQ-1384?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15902450#comment-15902450
]
Stanley Sung commented on HAWQ-1384:
------------------------------------
no one should bypass runaway detector. if resource queue defined and queries
are run as normal role, query should spill to disk without high vmem. So I
change this as minor.
> runaway_detector_activation_percent = 100 doesn't disable runaway_detector
> --------------------------------------------------------------------------
>
> Key: HAWQ-1384
> URL: https://issues.apache.org/jira/browse/HAWQ-1384
> Project: Apache HAWQ
> Issue Type: Bug
> Components: Query Execution
> Reporter: Stanley Sung
> Assignee: Lei Chang
> Priority: Minor
>
> When setting runaway_detector_activation_percent to 100, we still see queries
> canceled by runaway_detector like the following:
> ERROR: Canceling query because of high VMEM usage. Used: 64MB, available
> 81856MB, red zone: -16MB (runaway_cleaner.c:152) (seg9 xxxxx.xxxx.xxxx:40000
> pid=646507) (dispatcher.c:1801)
> Note that redzone is negative and used mem is way less than available.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
