felix-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Guillaume Nodet (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FELIX-2787) [File Install] Do not perform management activities while framework is starting/stopping
Date Sat, 17 Mar 2012 15:17:38 GMT

    [ https://issues.apache.org/jira/browse/FELIX-2787?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13231991#comment-13231991

Guillaume Nodet commented on FELIX-2787:

Would adding a check like
   context.getBundle(0).getState() == State.ACTIVE
 be sufficient ?

> [File Install] Do not perform management activities while framework is starting/stopping
> ----------------------------------------------------------------------------------------
>                 Key: FELIX-2787
>                 URL: https://issues.apache.org/jira/browse/FELIX-2787
>             Project: Felix
>          Issue Type: Improvement
>          Components: File Install
>    Affects Versions: fileinstall-3.1.4
>            Reporter: Richard S. Hall
>            Assignee: Guillaume Nodet
> File Install has been known to cause deadlocks, race conditions, and other sorts of spurious
issues. One of the main reasons for this is that File Install is pretty aggressive in its
management of bundles. This has caused us to improve the framework to deal with its aggressiveness,
but still it is not perfect. We have seen people wanting to introduce a delay value for management,
etc. We also see issues where the framework is trying to shut down and File Install is going
right behind the framework restarting bundles as the framework stops them. It would be better
if File Install monitored the starting/stopping status for the framework and only performed
its management activities while the framework were active. This means File Install's management
threads should not do their processing when the framework is not in the ACTIVE state. This
will be a good improvement, although it won't completely eliminate the window, since this
is a check-then-act situation. However, as long as the threads check the status on each process
loop, the window will be significantly reduced.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira


View raw message