[
https://issues.apache.org/jira/browse/IGNITE-4676?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15859219#comment-15859219
]
Alexey Kuznetsov edited comment on IGNITE-4676 at 2/9/17 9:27 AM:
------------------------------------------------------------------
Fixed. Test added.
Ready for review: http://reviews.ignite.apache.org/ignite/review/IGNT-CR-91
was (Author: kuaw26):
Fixed. Test added.
> Closure hangs if executed nested internal task with continuation
> ----------------------------------------------------------------
>
> Key: IGNITE-4676
> URL: https://issues.apache.org/jira/browse/IGNITE-4676
> Project: Ignite
> Issue Type: Bug
> Components: compute
> Affects Versions: 1.8
> Reporter: Alexey Kuznetsov
> Assignee: Alexey Kuznetsov
> Fix For: 1.9
>
>
> Initially this bug found in following scenario.
> Client send task T1 to node B via node A.
> Node A sends closure ExeCallable to node B and add listener to taskFuture.
> On node B ExeCallable executed internal task with one job with continuation.
> Result of ExeCallable never returned to node A.
> The problem isolated in GridJobWorker
> {code}private static final ThreadLocal<Boolean> HOLD{code}
> variable does not change its state in valid way for internal tasks and never
> call {{finishJob()}} method.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
