Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 339426 - Hudson is not 'letting go' of completed jobs
Summary: Hudson is not 'letting go' of completed jobs
Status: RESOLVED FIXED
Alias: None
Product: Community
Classification: Eclipse Foundation
Component: CI-Jenkins (show other bugs)
Version: unspecified   Edit
Hardware: PC Linux
: P3 normal (vote)
Target Milestone: ---   Edit
Assignee: Eclipse Webmaster CLA
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-03-09 16:08 EST by Denis Roy CLA
Modified: 2013-06-05 08:55 EDT (History)
7 users (show)

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Denis Roy CLA 2011-03-09 16:08:55 EST 
Jobs seem to complete, yet Hudson keeps them in the executor thread, tying up the server.
Comment 1 Denis Roy CLA 2011-03-09 16:09:17 EST 
Removing cross-project.
Comment 2 Denis Roy CLA 2011-03-09 16:50:18 EST 
The stalls seem to have started occurring recently.

Recent Hudson changes include the timeout plugin (and settings in some/all projects) and the Distributed Workspace Cleanup plugin (and its usage in some/all projects).
Comment 3 Denis Roy CLA 2011-03-09 17:00:35 EST 
With the workspace cleanup plugin enabled, is it possible that it's looking for hudson-slave2 (and stalling)?
Comment 5 Denis Roy CLA 2011-03-09 17:07:59 EST 
I've disabled the Distributed workspace plugin and restarted Hudson.  One job seems to be running now, and it's not stalled at start like the others were.
Comment 6 Denis Roy CLA 2011-03-10 08:39:48 EST 
Looks like the Distributed Workspace plugin was the culprit, as things seem to be running smoothly now.

We'll leave it disabled until after Indigo has shipped or until we run out of disk space, whichever comes first.
Comment 7 David Williams CLA 2011-03-10 08:47:09 EST 
(In reply to comment #6)
> Looks like the Distributed Workspace plugin was the culprit, as things seem to
> be running smoothly now.
> 
> We'll leave it disabled until after Indigo has shipped or until we run out of
> disk space, whichever comes first.

good catch. 
Should a bug be opened against hudson/plugin?