You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@impala.apache.org by "Gabor Kaszab (JIRA)" <ji...@apache.org> on 2017/11/29 14:06:00 UTC
[jira] [Resolved] (IMPALA-1144) Query cancellation throws error,
reports wrong query text
[ https://issues.apache.org/jira/browse/IMPALA-1144?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gabor Kaszab resolved IMPALA-1144.
----------------------------------
Resolution: Fixed
Fix Version/s: Impala 2.12.0
https://gerrit.cloudera.org/#/c/8549/
> Query cancellation throws error, reports wrong query text
> ---------------------------------------------------------
>
> Key: IMPALA-1144
> URL: https://issues.apache.org/jira/browse/IMPALA-1144
> Project: IMPALA
> Issue Type: Bug
> Components: Backend
> Affects Versions: Impala 1.4
> Environment: impala-shell, cancelling via ^C, while displaying a billion-row result set
> Reporter: John Russell
> Assignee: Gabor Kaszab
> Priority: Minor
> Labels: ramp-up, shell
> Fix For: Impala 2.12.0
>
>
> I was experimenting with a SELECT DISTINCT query for a unique column with a billion values, to see if it would exceed the mem limit. It eventually did start returning results. When I hit ^C to cancel in impala-shell, the result looked like this:
> | 366727153 |
> | 133154612 |
> | 596104894 |
> | 242600935 |
> | 474071842 |
> | 307379041 |
> | 604906030 |
> | 61000172 |
> | 621687629 |
> +-----------+
> ERROR: Invalid query handle
> Query: use oreilly
> I think impala-shell should handle the cancellation gracefully rather than printing an "invalid query handle" error. What's really confusing is why it echoed
> Query: use oreilly
> which was the statement that immediately preceded my SELECT query. (The USE was run behind the scenes by 'impala-shell -d oreilly'.) It looks like impala-shell forgot which query it was cancelling.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)