Has anyone ever seen this problem?

Fri Oct 21 16:05:31 MST 2005

Is the output posted in your mail a grep of the messages, or the actual
file?

The reason I ask is that there is no information in your post about actual
reboots, in which case of course the typical system startup messages would
be present.

In the past I have seen the 'ocassional' problem where syslog itself is
randomly restarting.

Also I had a client a few years back who believed they had similar server
issues, their symptoms were that the Server would become unreachable every
few hours. After checking the logs I saw continuous restart messages like
yours, however there were no new dmesgs. I then payed a visit to the site
and found that they had taken it upon themselves to move the server and
switch into a closet.

I suspected there was a problem with heat dissipation, and wrote a quick
script to ping the workstations every 5 minutes. This showed network
connectivity dropping at approximately every syslog restart message time.
But there was still no other output in the logs on the Server. I called them
the next day and had them leave the closet door open. No loss of
connectivity occured that day.

The switch was overheating. I never determined why I was getting the restart
message in the syslogs because they stopped appearing in the logs after
that. I can only guess that there was a bug at work with the tcp stack
interacting with syslog and causing the restart message.

On 10/21/05, der.hans <PLUGd at lufthans.com> wrote:
>
> Am 21. Oct, 2005 schwätzte JD Austin so:
>
> > I have a new client that has a weird problem with their linux file
> server.
> > It is rebooting without warning several times a day.
>
> You might want to start grabbing system info to see what's going on prior
> to the shutdowns.
>
> I have a script called sys_stats.sh you could look at. I haven't in a long
> time, so it might be out of date...
>
> https://www.LuftHans.com/unix/
>
> My script doesn't capture temperature info. gkrellm will do that for you.
> I think it has a 'save the info' option.
>
> I believe there's a debian package that has tools for stuff like my
> sys_stats script. Probably better than my script, but I didn't see the
> package with a quick search.
>
> Turn up the syslog output and capture at the lowest level, debug.
>
> Is there anything else plugged in to that UPS?
>
> When plugged in to the wall was the UPS unplugged?
>
> Causes I can think of:
>
> not enough juice ( the 500 W power supply should take care of that )
> overheating ( the extra fans should help, but maybe cpu fan is on funny )
> buggy mboard or ram ( install one of the memtest packages and test )
> buggy UPS
> someone keeps rebooting it ( leave the keyboard there, but unattached )
>
> The last one should be causing syslog output.
>
> ciao,
>
> der.hans
>
> > reboot activity
> > *Oct 20 06:26:45 aflsdebprod syslogd 1.4.1#10: restart.*
> > Oct 20 08:48:40 aflsdebprod syslogd 1.4.1#10: restart.
> > *Oct 20 11:45:48 aflsdebprod syslogd 1.4.1#10: restart.*
> > Oct 20 11:52:02 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 11:58:02 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 12:07:50 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 12:15:38 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 12:45:41 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 13:15:42 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 13:24:20 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 13:38:38 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 14:04:08 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 14:15:13 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 14:37:59 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 15:10:02 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 15:43:59 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 16:00:47 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 17:13:55 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 17:22:14 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 17:36:42 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 17:41:52 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 18:00:38 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 18:30:41 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 19:00:38 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 20:00:39 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 20:05:15 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 20 22:23:12 aflsdebprod syslogd 1.4.1#10: restart.
> > *Oct 21 06:26:45 aflsdebprod syslogd 1.4.1#10: restart.
> > Oct 21 11:45:59 aflsdebprod syslogd 1.4.1#10: restart.
> >
> > Notice how it happened almost at the exact same time yesterday.
> >
> > *Since it only had a 330W power supply and several extra fans in the
> case
> > I replaced the power supply with a 500W power supply last night.
> > The computer is plugged into a UPS. Yesterday prior to getting a new
> > power supply I had them plug the machine directly into the wall to see
> if
> > it was the UPS causing it, the problem persisted.
> >
> >
> > I can find nothing in the logs that would indicate that hardware
> > is failing, they all look like this:
> >
> > Oct 21 04:23:11 aflsdebprod -- MARK -- Oct 21 11:45:59 aflsdebprod
> syslogd
> > 1.4.1#10: restart.
> > Oct 21 11:45:59 aflsdebprod kernel: klogd 1.4.1#10, log source =
> /proc/kmsg
> > started.
> > Oct 21 11:45:59 aflsdebprod kernel: Inspecting /boot/System.map-
> 2.4.18-bf2.4
> >
> >
> > Anyone had this happen to them before? JD
> >
> >
>
> --
> # https://www.LuftHans.com/
> # If you have an apple and I have an apple and we exchange apples then
> # you and I will still each have one apple. But if you have an idea and
> # I have an idea and we exchange these ideas, then each of us will have
> # two ideas. -- George Bernard Shaw
>
> ---------------------------------------------------
> PLUG-discuss mailing list - PLUG-discuss at lists.plug.phoenix.az.us
> To subscribe, unsubscribe, or to change you mail settings:
> http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.plug.phoenix.az.us/pipermail/plug-discuss/attachments/20051021/573a1310/attachment-0001.htm