Bugzilla – Full Text Bug Listing
| Summary: | Apache failing to hup correctly after logrotate | ||
|---|---|---|---|
| Product: | Community | Reporter: | Denis Roy <denis.roy> |
| Component: | Servers | Assignee: | Eclipse Webmaster <webmaster> |
| Status: | RESOLVED WORKSFORME | QA Contact: | |
| Severity: | blocker | ||
| Priority: | P3 | CC: | gunnar, karl.matthias, kim.moir, richard.gronback, wgp010 |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Linux | ||
| Whiteboard: | |||
|
Description
Denis Roy
*** Bug 256419 has been marked as a duplicate of this bug. *** *** Bug 256196 has been marked as a duplicate of this bug. *** This appears to be a PHP bug. By the looks of the bug report, chances of it being fixed are slim to none: http://bugs.php.net/bug.php?id=44309 on dsdp mtj svn the following error is shown now: Some of selected resources were not committed. svn: Commit failed (details follow): svn: Can't open file '/home/data/svn/dsdp/org.eclipse.mtj/db/write-lock': Permission denied svn: MERGE of '/svnroot/dsdp/org.eclipse.mtj/trunk/releng/org.eclipse.mtj.releng': 409 Conflict (https://dev.eclipse.org) Denis, Matt and I also found this when you were out of town and three of the nodes crashed at various times. We also tracked it down to a PHP bug at that time. Pretty much sucks. I remember vividly because with backoffs Nagios wouldn't page until about 4:30am my time, just about right for ruining a night of sleep. ;) I really thought we had discussed it when you got back. Sorry if we didn't. Make that 1:30. :) We had discussed this, but the problem magically went away, so I simply didn't believe you :) I do now. Until we can solve this, rasputin will kill -9 the httpd processes and restart them. It works manually. So after the last time this happened I added a 'killproc' option for the apache process to try and force a restart. However a little while later I added some extra code to skip specific service checks when those services had reached their 'flap' limit. Catch is < is not > and so once the service had failed once it was skipping new failures and failing to restart. I've pulled that code and restarted rasputin. Which may also explain the strange 'reappearance' of this issue. -M. i still can't commit on MTJ SVN. is it still the same issue? the error is on comment 4 (In reply to comment #8) > So after the last time this happened I added a 'killproc' option for the apache > process to try and force a restart. However a little while later I added some > extra code to skip specific service checks when those services had reached > their 'flap' limit. Catch is < is not > and so once the service had failed > once it was skipping new failures and failing to restart. I've pulled that > code and restarted rasputin. Which may also explain the strange 'reappearance' > of this issue. > > -M. > (In reply to comment #9) > i still can't commit on MTJ SVN. is it still the same issue? the error is on > comment 4 The issue with SVN is not directly connected with the issue with apache. The SVN issue can be tracked on bug 256436. -M. Closing as we haven't seen this issue after the last comment. -M. Moving all these to Servers. |