You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Khurram Faraaz (JIRA)" <ji...@apache.org> on 2017/08/30 11:38:00 UTC
[jira] [Commented] (DRILL-4296) Query hangs in
CANCELLATION_REQUESTED when cancelled after it starts returning results
[ https://issues.apache.org/jira/browse/DRILL-4296?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16147097#comment-16147097 ]
Khurram Faraaz commented on DRILL-4296:
---------------------------------------
[~kkhatua] Executed same query over same data from this JIRA on Drill 1.12.0-SNAPSHOT commit a51c98b8bf210bbe9d3f4018361d937252d1226d. Canceled query from sqlline and query moves to CANCELED state. Query does not remain or move to CANCELLATION_REQUESTED state once it is canceled from sqlline.
> Query hangs in CANCELLATION_REQUESTED when cancelled after it starts returning results
> --------------------------------------------------------------------------------------
>
> Key: DRILL-4296
> URL: https://issues.apache.org/jira/browse/DRILL-4296
> Project: Apache Drill
> Issue Type: Bug
> Components: Execution - Flow
> Affects Versions: 1.5.0
> Environment: commit.id=c9dbfbd
> 2 nodes with 32 cores and 32GB of max direct memory for drill
> Reporter: Deneche A. Hakim
> Attachments: 295eefc3-8d15-d63b-a721-3fde365b639c.sys.drill, data.tar.gz, node1_jstack.txt, node2_jstack.txt
>
>
> After running the following queries (it's the same reproduction from DRILL-2274):
> {noformat}
> set planner.memory.max_query_memory_per_node=8589934592;
> select sub1.uid from `all2274.json` sub1 inner join `all2274.json` sub2 on sub1.uid = sub2.uid order by sub1.uid;
> {noformat}
> After the query starts returning results, I cancelled the query from sqlline. This caused the query to hang in a CANCELLATION_REQUESTED state.
> Looking at jstack (attached) the root fragment is blocked waiting for Ack from the client.
> The the foreman node (which also runs Zookeeper) runs out of disk space once the query finishes spilling, which seems to contribute to the issue. Once I changed the spill directory to nfs I no longer so the issue.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)