[
https://issues.apache.org/jira/browse/DRILL-3640?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16061699#comment-16061699
]
ASF GitHub Bot commented on DRILL-3640:
---------------------------------------
Github user laurentgo commented on a diff in the pull request:
https://github.com/apache/drill/pull/858#discussion_r123865940
--- Diff:
exec/jdbc/src/main/java/org/apache/drill/jdbc/impl/DrillStatementImpl.java ---
@@ -38,8 +44,12 @@
// methods for compatibility.)
class DrillStatementImpl extends AvaticaStatement implements
DrillStatement,
DrillRemoteStatement {
+ //Not using the DrillbitContext's ExecutorService as this is threadPool
is light-weight (threads wake up to cancel tasks) but needs a low response time
+ private static ExecutorService queryTimeoutTaskPool =
Executors.newCachedThreadPool(new NamedThreadFactory("q-timeout-"));
--- End diff --
Maybe simpler, but I'm doubtful this is less overhead: there's a new
executor and a new thread created for each statement (to compare with just
telling the system how long to wait in DrillCursor, knowing that the i/o system
is using a fixed number of threads vs your proposal which is unbounded).
I'm also not following why you would need a new state, as this would be a
pure client thing.
> Drill JDBC driver support Statement.setQueryTimeout(int)
> --------------------------------------------------------
>
> Key: DRILL-3640
> URL: https://issues.apache.org/jira/browse/DRILL-3640
> Project: Apache Drill
> Issue Type: New Feature
> Components: Client - JDBC
> Affects Versions: 1.2.0
> Reporter: Chun Chang
> Assignee: Kunal Khatua
> Fix For: 1.11.0
>
>
> It would be nice if we have this implemented. Run away queries can be
> automatically canceled by setting the timeout.
> java.sql.SQLFeatureNotSupportedException: Setting network timeout is not
> supported.
> at
> org.apache.drill.jdbc.impl.DrillStatementImpl.setQueryTimeout(DrillStatementImpl.java:152)
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
