PFSense - Reboot Randomly

[wisowebs]

Hey All:

I have had a PFSense box in and running along smooth in a VM environment for close to 4 months, it has been super solid.  Lately we have put a monitoring solution in and have found the firewall randomly rebooting itself after hours.  This is the only machine doing this, and the logs are rather greek to me.  Just curious is anyone has any similar issues, the only thing changed in the last few days is enabling SNMP.  Here is a snipped from the logs when it appeared to go down, however this may very well be when it was on its way up.  Any insight I could get would be great…

Jul 10 00:39:40 kernel: current process = 12 (swi6: task queue)
Jul 10 00:39:40 kernel: processor eflags = interrupt enabled, resume, IOPL = 0
Jul 10 00:39:40 kernel: = DPL 0, pres 1, long 1, def32 0, gran 1
Jul 10 00:39:40 kernel: code segment = base 0x0, limit 0xfffff, type 0x1b
Jul 10 00:39:40 kernel: frame pointer = 0x28:0xffffff800008ab00
Jul 10 00:39:40 kernel: stack pointer = 0x28:0xffffff800008aae0
Jul 10 00:39:40 kernel: instruction pointer = 0x20:0xffffffff806ee7fe
Jul 10 00:39:40 kernel: fault code = supervisor read data, page not present
Jul 10 00:39:40 kernel: fault virtual address = 0x290
Jul 10 00:39:40 kernel: cpuid = 0; apic id = 00
Jul 10 00:39:40 kernel: Fatal trap 12: page fault while in kernel mode
Jul 10 00:39:40 kernel:
Jul 10 00:39:40 kernel:
Jul 10 00:39:40 kernel:
Jul 10 00:39:40 kernel: acd0: WARNING - PREVENT_ALLOW freeing taskqueue zombie requestWARNING - PREVENT_ALLOW freeing taskqueue zombie request
Jul 10 00:39:40 kernel: acd0:
Jul 10 00:39:40 kernel: acd0: acd0: WARNING - PREVENT_ALLOW taskqueue timeout - completing request directlyWARNING - PREVENT_ALLOW taskqueue timeout - completing request directly
Jul 10 00:39:40 syslogd: kernel boot file is /boot/kernel/kernel

Any ideas, not mission critical but if it took a reboot during business hours we would feel it..?

[kejianshi]

I can't take any credit for diagnosing this.
My first thoughts was bad hardware, but seems lots of you are getting this after a recent update.

http://forum.pfsense.org/index.php/topic,64144.0/topicseen.html

[wisowebs]

Thanks for your replay, my version is

2.0.3-RELEASE (amd64)
built on Fri Apr 12 10:27:15 EDT 2013

Nothing has changed in our hardware, it actually is rock solid running quite a bit of other vms.  Ubuntu, Windows, Free BSD, etc in our VMWare environment.  We have a large amount of virtual machines in our cluster so if we had hardware or VmWare issues we would know about it for sure, this has to be proprietary to PFSense in some way shape or form, just not sure what yet.

[kejianshi]

I'm not sure if this applies to you in your environment but, depending on the packages you run, it very well might.

http://doc.pfsense.org/index.php/Tuning_and_Troubleshooting_Network_Cards

[wisowebs]

Thank, but yeah that does not apply to our hardware what-so-ever.

[wallabybob]

@wisowebs:

Jul 10 00:39:40 kernel: acd0: WARNING - PREVENT_ALLOW freeing taskqueue zombie requestWARNING - PREVENT_ALLOW freeing taskqueue zombie request
Jul 10 00:39:40 kernel: acd0:
Jul 10 00:39:40 kernel: acd0: acd0: WARNING - PREVENT_ALLOW taskqueue timeout - completing request directlyWARNING - PREVENT_ALLOW taskqueue timeout - completing request directly
Jul 10 00:39:40 syslogd: kernel boot file is /boot/kernel/kernel

It looks as if the kernel is attempting to write a message about a CD-ROM device acd0. Does the pfSense VM have an emulated CD or access a real CD? If the CD is not needed I would remove it from the VM to see if that makes a difference.

[wisowebs]

Thanks wallabybob…

Shows different than acd0 in VMWare but makes sense, will be trying that, will let you know how it works.

Thanks for everyones responses!!!