Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 7 Jul 2002 18:14:34 -0400 (EDT)
From:      Kenneth W Cochran <kwc@TheWorld.com>
To:        freebsd-stable@freebsd.org
Subject:   Out of swap, -stable, apache2
Message-ID:  <200207072214.SAA4562252@shell.TheWorld.com>

next in thread | raw e-mail | index | archive | help
Hello -stable:

Is anyone else here having an out-of-swap condition, apparently
caused by apache2?

This has only been happening since about the end of June.
It seems to start about 24 hours after the system has been
rebooted after a cvsup/{build,install}world.

OS is -stable, as of 2002/07/05; this started happening (I
think) after some port-updating (apache2) on 28 June.
Webserver is apache2, from -ports, along with mod_php4.
 Apache/2.0.39 (Unix) DAV/2 PHP/4.2.1

While this event is underway, top shows httpd at the top
(eating quite a lot of CPU) & with a status of "PFAULT"
(or something like that).  After the event "finishes" httpd &
the system appear to be running again, just fine.

Looking around in the logs, I find the following:

From /var/log/messages:

Jul  6 19:42:09 <kern.crit> localhost /kernel: swap_pager: out of swap space
Jul  6 19:42:09 <kern.crit> localhost /kernel: swap_pager_getswapspace: failed
Jul  6 19:42:09 <kern.err> localhost /kernel: pid 33475 (httpd), uid 80, was killed: out of swap space
.
.
Jul  6 22:36:39 <kern.crit> localhost /kernel: swap_pager_getswapspace: failed
Jul  6 22:36:39 <kern.crit> localhost last message repeated 46 times
Jul  6 22:36:39 <kern.err> localhost /kernel: pid 33478 (httpd), uid 80, was killed: out of swap space
.
.
Jul  7 11:25:41 <kern.crit> localhost /kernel: swap_pager_getswapspace: failed
Jul  7 11:25:41 <kern.crit> localhost last message repeated 64 times
Jul  7 11:25:41 <kern.err> localhost /kernel: pid 46115 (httpd), uid 80, was killed: out of swap space

From /var/log/httpd-error.log:  (ip-addresses clobbered to
    hopefully not embarrass the guilty... :)

[Sat Jul 06 19:40:32 2002] [error] [client 999.999.99.999] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
.
.
[Sat Jul 06 22:35:19 2002] [error] [client 888.888.888.88] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /
.
.
[Sun Jul 07 11:24:16 2002] [error] [client 777.77.777.777] client sent HTTP/1.1 request without hostname (see RFC2616 section 14.23): /

Anything else in the logs is several minutes "outside" these
(perhaps related?) entries.

DNS is cache-only:
named 8.3.2-T1B Fri Jul  5 19:08:16 CDT 2002
        root@localhost.my.domain:/usr/obj/usr/src/usr.sbin/named

Any ideas as to what might be going on here?

The above is, to me, admittedly rather sketchy; any suggestions
as to how I might trigger this myself & capture any more helpful
logging/debugging data/info?

Is this even a FreeBSD issue?  i.e. Should I take this up with
the Apache Group?  Or is this some kind of pilot-error on my part?

Many thanks,

-kc

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200207072214.SAA4562252>