You are viewing a plain text version of this content. The canonical link for it is here.
Posted to apache-bugdb@apache.org by Stuart Peters <st...@mill.co.uk> on 2001/11/22 16:34:09 UTC

general/8836: Apache grabbed disk space in /tmp but data not visible through OS

>Number:         8836
>Category:       general
>Synopsis:       Apache grabbed disk space in /tmp but data not visible through OS
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    apache
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   apache
>Arrival-Date:   Thu Nov 22 07:40:00 PST 2001
>Closed-Date:
>Last-Modified:
>Originator:     stuart@mill.co.uk
>Release:        1.3.20
>Organization:
apache
>Environment:
Linux beam.mill.co.uk 2.4.2-2 #1 Sun Apr 8 20:41:30 EDT 2001 i686 unknown

Linux system running RedHat Seawolf, server compiled on a different but identical server
>Description:
/tmp went 100% full and server stopped taking file uploads.  However, /tmp showed only 200Mb data in the file system, out of a 1Gb partition - 800Mb had 'disappeared'.  An Apache child process was holding onto /tmp so we couldn't unmount to check disk, restarting Apache worked fine and cleared the problem, releasing the 800Mb data.

The only unusual errors logged that day were:

[Tue Nov 20 10:40:11 2001] [warn] child process 30568 still did not exit, sending a SIGTERM
[Tue Nov 20 10:40:11 2001] [warn] child process 30569 still did not exit, sending a SIGTERM
[Tue Nov 20 10:40:15 2001] [notice] caught SIGTERM, shutting down

...but it is more likely that the problem only began at around 7pm of that evening.
>How-To-Repeat:
Unfortunately I don't understand why this happened.  My only guess is that incoming data exceeded the 800Mb of free space and caused some failure, but this is speculation.  An attempted upload of this size to our server would be unusual but not impossible.
>Fix:
Restarting Apache released the 'held' file space.

There is no certainty that this was caused by a bug in Apache, but it was an Apache child process that appeared to be holding the disk space in /tmp.
>Release-Note:
>Audit-Trail:
>Unformatted:
 [In order for any reply to be added to the PR database, you need]
 [to include <ap...@Apache.Org> in the Cc line and make sure the]
 [subject line starts with the report component and number, with ]
 [or without any 'Re:' prefixes (such as "general/1098:" or      ]
 ["Re: general/1098:").  If the subject doesn't match this       ]
 [pattern, your message will be misfiled and ignored.  The       ]
 ["apbugs" address is not added to the Cc line of messages from  ]
 [the database automatically because of the potential for mail   ]
 [loops.  If you do not include this Cc, your reply may be ig-   ]
 [nored unless you are responding to an explicit request from a  ]
 [developer.  Reply only with text; DO NOT SEND ATTACHMENTS!     ]