[jira] [Resolved] (IGNITE-1044) HadoopJob#dispose(bool) should be called for each HadoopJob instance in any case

["Ivan Veselovsky (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IGNITE-1044?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ivan Veselovsky resolved IGNITE-1044.
-------------------------------------
    Resolution: Cannot Reproduce

Unfortunately, there  is no simple way to guarantee #dispose() to be called in any case for the Job object -- the current design does not allow that.

Running entire the Hadoop test suite with special probes that must show not disposed Jobs does not reveal the problematic situation. (Not disposed Jobs occur in some synthetic tests and in some rare cases when the tests terminate ubnormally, e.g. upon a timeout).

> HadoopJob#dispose(bool) should be called for each HadoopJob instance in any case
> --------------------------------------------------------------------------------
>
>                 Key: IGNITE-1044
>                 URL: https://issues.apache.org/jira/browse/IGNITE-1044
>             Project: Ignite
>          Issue Type: Bug
>          Components: hadoop
>    Affects Versions: sprint-6
>            Reporter: Ivan Veselovsky
>            Assignee: Ivan Veselovsky
>             Fix For: 1.8
>
>
> Now there are some rare cases when for some HadoopJob instances this method is not called. This sometimes observed in tests. 
> It looks like these cases are not related to any failures.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)