[ https://issues.apache.org/jira/browse/IMPALA-12951?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17843919#comment-17843919 ]
Michael Smith commented on IMPALA-12951: ---------------------------------------- Impala has a general query retry mechanism. It's not used for inserts. I believe this problem only applies to system tables, but not 100% sure. > Investigate RETRY_FAILED_QUERIES with System Tables > --------------------------------------------------- > > Key: IMPALA-12951 > URL: https://issues.apache.org/jira/browse/IMPALA-12951 > Project: IMPALA > Issue Type: Task > Reporter: Michael Smith > Priority: Major > > My theory is that {{RETRY_FAILED_QUERIES}} doesn't work well with querying > System Tables because coordinators don't get added to a blacklist when they > fail, so it reschedules them on the same set of nodes. > Look into whether issues with {{RETRY_FAILED_QUERIES}} when a coordinator > restarts (without graceful shutdown) are unique to coordinators, and how to > improve it. -- This message was sent by Atlassian Jira (v8.20.10#820010) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-all-unsubscr...@impala.apache.org For additional commands, e-mail: issues-all-h...@impala.apache.org