Re: Why does MaintainSwapSpace run continuously when swap is full?

This is the mail from Duane, it could be the bug that its bitting you in
2.3Stable3:

Hi All,

The Squid-2.3.STABLE3 release contains a serious bug that causes your
disks to fill up. If you're using this version, please visit the bugs
page and download the patch for the "disk space over the limit" bug.

http://www.squid-cache.org/Versions/v2/2.3/bugs/

Duane W.

----- Original Message -----
From: <wesf@us.ibm.com>
To: <squid-users@ircache.net>
Sent: Friday, May 26, 2000 10:11 AM
Subject: Why does MaintainSwapSpace run continuously when swap is full?

>
>
> I am running Web Polygraph against Squid (although I don't think this is a
> Poly-specific problem), and I notice that as soon as the swap fills up,
the
> MaintainSwapSpace task runs, but since nothing in the cache is expired it
> doesn't remove any objects. Since the cache is still full, it reschedules
> itself to run immediately, consuming all available CPU time. Squid
> continues to handle requests during this time, but I don't understand why
> MaintainSwapSpace has this behavior. Is this a bug or a misconfiguration?
>
> This happens with both Squid 2.3-STABLE3 and 2.4-DEVEL2 on Red Hat 6.2.
>
> Here's an excerpt from the cache.log while this is happening:
>
> 2000/05/26 09:47:44| eventRun: RUN ID 1269789
> 2000/05/26 09:47:44| eventRun: Running 'MaintainSwapSpace', id 1269788
> 2000/05/26 09:47:44| eventAdd: Adding 'MaintainSwapSpace', in 0.000000
> seconds
> 2000/05/26 09:47:44| storeMaintainSwapSpace: f=1.000000, max_scan=500,
> max_remove=80
> 2000/05/26 09:47:44| storeMaintainSwapSpace: scanned 500/500 removed 0/80
> locked 0 f=1.000
> 2000/05/26 09:47:44| storeMaintainSwapSpace stats:
> 2000/05/26 09:47:44| 156786 objects
> 2000/05/26 09:47:44| 500 were scanned
> 2000/05/26 09:47:44| 0 were locked
> 2000/05/26 09:47:44| 0 were expired
> 2000/05/26 09:47:44| comm_poll: 0 FDs ready
> 2000/05/26 09:47:44| comm_poll: time out: 959352464.
> 2000/05/26 09:47:44| eventRun: RUN ID 1269790
> 2000/05/26 09:47:44| eventRun: Running 'MaintainSwapSpace', id 1269789
> 2000/05/26 09:47:44| eventAdd: Adding 'MaintainSwapSpace', in 0.000000
> seconds
> 2000/05/26 09:47:44| storeMaintainSwapSpace: f=1.000000, max_scan=500,
> max_remove=80
> 2000/05/26 09:47:44| storeMaintainSwapSpace: scanned 500/500 removed 0/80
> locked 0 f=1.000
> 2000/05/26 09:47:44| storeMaintainSwapSpace stats:
> 2000/05/26 09:47:44| 156786 objects
> 2000/05/26 09:47:44| 500 were scanned
> 2000/05/26 09:47:44| 0 were locked
> 2000/05/26 09:47:44| 0 were expired
> 2000/05/26 09:47:44| comm_poll: 0 FDs ready
> 2000/05/26 09:47:44| comm_poll: time out: 959352464.
> 2000/05/26 09:47:44| eventRun: RUN ID 1269791
> 2000/05/26 09:47:44| eventRun: Running 'MaintainSwapSpace', id 1269790
> 2000/05/26 09:47:44| eventAdd: Adding 'MaintainSwapSpace', in 0.000000
> seconds
> 2000/05/26 09:47:44| storeMaintainSwapSpace: f=1.000000, max_scan=500,
> max_remove=80
> 2000/05/26 09:47:44| storeMaintainSwapSpace: scanned 500/500 removed 0/80
> locked 0 f=
> 2000/05/26 09:47:48| WARNING: Disk space over limit: 2188830 KB > 1048576
> KB
> 2000/05/26 09:47:58| WARNING: Disk space over limit: 2188830 KB > 1048576
> KB
> 2000/05/26 09:48:08| WARNING: Disk space over limit: 2188830 KB > 1048576
> KB
>
> Wesley Felter - wesf@us.ibm.com
>
Received on Sat May 27 2000 - 08:45:10 MDT