[ https://issues.apache.org/jira/browse/HAWQ-81?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14975350#comment-14975350 ]
ASF GitHub Bot commented on HAWQ-81: ------------------------------------ GitHub user jiny2 opened a pull request: https://github.com/apache/incubator-hawq/pull/43 HAWQ-81. Make resource manager to reject resource request when there … …are too many segments unavailable. This improvement makes HAWQ RM able to reject resource request directly if there are too many segments unavailable. The expect cluster size is read from $GPHOME/etc/slaves. The new guc name is hawq_rm_rejectrequest_nseg_limit. The default value is 2, i.e. if there are more than 2 segments unavailable, HAWQ RM rejects the request directly. The reject error message looks as below. postgres=# insert into t values(1); ERROR: failed to acquire resource because of too many unavailable segments. (pquery.c:798) You can merge this pull request into a Git repository by running: $ git pull https://github.com/jiny2/incubator-hawq HAWQ-81 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/incubator-hawq/pull/43.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #43 ---- commit b8678fd20197ceeb43045adbc12d0079bcadf42a Author: Yi Jin <y...@pivotal.io> Date: 2015-10-26T23:33:15Z HAWQ-81. Make resource manager to reject resource request when there are too many segments unavailable. ---- > Make resource manager to reject resource request when there are too many > segments down > -------------------------------------------------------------------------------------- > > Key: HAWQ-81 > URL: https://issues.apache.org/jira/browse/HAWQ-81 > Project: Apache HAWQ > Issue Type: Improvement > Components: Resource Manager > Reporter: Yi Jin > Assignee: Yi Jin > > If there are too many physical segments down, the performance of queries > maybe seriously impacted, continue processing queires does not make sense > sometime. Therefore, new guc hawq_rm_rejectrequest_nseg_limit is introduced, > default value 2, minimum value 0. > User can configure this value to let resource manager reject resource request > immediately when more than hawq_rm_reject_nseg_down_limit segments are down. > The expected full cluster size ( size of segments ) is read from file > $GPHOME/etc/slaves. -- This message was sent by Atlassian JIRA (v6.3.4#6332)