Running FreeBSD 5.3-p10

I got a new server from TP.com (Total Control). It's a Dual P Xeon 2.8 with 1gb ram. Well every day (seems to be when I am not active like sleeping or at work), it freezes up. I had a tech go check it out and he said there was no errors on the screen and the console was frozen (not responding to the keyboard). I checked crontab and there is nothing running that may crash it I dont think (I'll post it below).

There is nothing running that is not necessary to my usual server services. I had a SM P4 2.8ghz with the same crontab and a few other configs (such as logging and apache) and it didn't freeze up ever.

Here is my crontab:



If anyone has any ideas, please post them here. I want this resolved asap please!

Also, I had a tech also test all the hardware, they swapped out the ram and this is still happening.

edit: also, it doesnt seem to be happening at any particular time (except while i'm asleep or at work). i've setup timestamp'd pings to the server and it stops responding at random times.

If the console is frozen, I'd say hardware, but you've already had them check that out?

Very strange.

Kind of an off shot but are you running a custom kernel? Only reason I ask is because one of my systems at home was locking up in the same way you describe while I was benchmarking SCHED_ULE with SMP on 5.3-STABLE about a month ago.

yeah it's custom.

options SCHED_4BSD # 4BSD scheduler

that? or something else?

Something else, you should be fine with that. I don't think it is cron related, looks like a default crontab with everything set to run in the middle of the night.

Are all your installed ports up to date? I'm not aware of any lockup issues with 5.3, to confirm it is not kernel related you could always try running a GENERIC kernel with SMP enabled. Also, try doing some heavy load operations to see if you can cause it to hang up.

Beyond that I can only suggest checking the logs for all your running services, or setting them to log as verbose as possible if they aren't already. Maybe try eliminating what services you can one by one to see if you can isolate the problem.

all ports are up to date. i cvsup'd all before i configured my kernel and such, like i said it's a new server too.

no logs suggest anything out of the ordinary

i think it might have something to do with bind maybe.

i was messing around setting up awstats and was restarting named. it was complaining about not having permissions to named.pid so i shanged ownsership to bind:bind. i started named, all was fine for under a minute then it locked up. nothing on the console, nothing in /var/log/messages

what else can i do to explore this?

MAybe try updating to the latest stable release?

i am running this same release on another server and it runs perfectly. so i dont understand what is wrong.

5.4-RELEASE is out, see if it fixes you issues. Also if you are using SCHED_ULE a lot of changes have been made in 5.4 so that could resolve the issues.

Besides that, i have not heard of any problems with 5.3-p10, so it could be hardware related.

it has restarted itself 2 times in 12 hours now. i will try going to 5.4 release.

a tech at ThePlanet logged in and checked around and also found nothing wrong.

oh my bad, it just went down a couple minutes ago, looks like TP is attempting a chassis swap. they were supposed to do it last night, but i guess they decided to pull it now and not even tell me. lol

chassis swap is done. no freezing up since it went back up, although it's too soon to tell so i will let everyone know what happens.

Yes please let us know...

still freezing up. *sigh*

cvs'd to 5.3-RELEASE and still happening as well. TP looks like they are attempting to help me for free (at least they say so lol).

i'll keep this updated.

Damn sorry to hear that. This problem has me baffled as well. Keep us updated if anything new comes up.

I had 2 5.3-R Celeron servers that were crashing every 2-3 days some time ago. I finally had to downgrade to 5.2.1-R and they are not crashing anymore. You may try this route. You can downgrade on FreeBSD.

Still have not heard if you have upgraded to 5.4-RELEASE yet, i would like to see what happens then. 5.4-RELEASE is extremely stable on my servers.

I'm having a similar problem. I'm running 5.4 Release and my server will just hang every now and again. I am logging everything I can find and usually the last line is one of the common cron jobs (the last one was with save-entropy). I'm running mysql, apache, postfix, vm-pop3d, php5, pop-before-smtp, sympa.. Nothing unusual. I am not running a custom kernal to my knowledge (unless the folks at aPlus set it up with one).

I'm kinda new to FreeBSD so I'm not exactly sure I'm doing everything I can to track down the problem. Is there a place where I can make the logs more robust? I have the all.log running and it didn't show anything. Fedora has a boot log that had all of its boot problems show up in, does FreeBSD have something similar? This is really driving me nuts.... I can't seem to find any reason at all for this box to be locking up.

Has there been anything more on this issue? Any help would greatly be appreciated!!!!

Sal.