Shutdown / reboot problem - NMI ... going to debugger

[Silvan Burch]

Hey guys

I'm pretty new to freebsd/freenas but I have some experience with Linux at least.
So I installed freenas 9.10 (first stable / after problems occurred updated to nightlies as you'll read) on my ASRock Rack C236 WSI with Intel i3-6100. Pretty new hardware but the freenas update site states it should be supported.

So I played around a bit, installed plugins and jail (e.g. sab, sickbeard, headphones, couchpotato and in jails openvpn and logitech media server).

All those plugins and jails work like a charm. Problem I have occurs during reboot or shutdown:
sometimes while syncing disks instead of:

"Syncing disks, vnodes remaining... 0 0 0 0 0"

the following shows

"Syncing disks, vnodes remaining... 0 NMI ISA 30, EISA FF" & "NMI ... going to debugger"

after that, the output goes crazy with a

"Tracing command kernel pid 0 tid XYZ"

where where XYZ is an increasing integer (picture's of both screens attached)

anyway, at this point it just remains until I manually power off the PC ...

To be honest, I have no idea what the problem could be ... I was not able to figure out a pattern for when that problem occurs ... sometimes it start and shuts down/reboot without a problem, sometimes not ...

I'd really appreciate any help on that.

Thanks and cheers, Silvan

 

[jgreco]

Please file a bug report.

 

[Silvan Burch]

bug reported:

https://bugs.freenas.org/issues/15089

 

[Silvan Burch]

also, memtest86 is running at the moment. the first pass showed no problem with the RAM (except that it does not show the right frequency - instead of 2133 MHz it shows 1067 MHz ... so half ...)
anyway, I think RAM should not be a problem

 

[jgreco]

Just so you're aware, a single pass of memtest86 means very little. You really need to run it for at least a week to be reasonably confident in your memory.

 

[Silvan Burch]

thanks for the input. It's still running atm so I have at least some reliable data.
I was just thinking that a RAM problem might lead to the problem as it occurs only sometimes ... so even if there is something wrong on these RAM, it is at least not a really bad defect (as it was on the last RAM I bought for my PC)

 

[jgreco]

No, the shutdown thing sounds like something else going awry.

 

[Silvan Burch]

Yeah I kinda hope RAM is ok :)

Anyway, thanks for helping me!

 

[cyberjock]

I would look at:

BIOS update if any exist.
Try factory defaulting the BIOS, then only changing the settings you *must* change to make the machine bootable.

 

[Silvan Burch]

hey cyberjock

sadly, there's no bios update yet ... also, I did not change anything in the bios ... it seemed to everythings in order the first time I started it

 

[Silvan Burch]

well ... I updated the bug report pretty often and tried different stuff that struck me as strange but nothing helped so far ...

if anyone has any idea? I'd really appreciate it ... thanks

 

[Silvan Burch]

well it seems I've found the problem ... kinda?
disabling all jails, all plugins AND removing static routes seems to do the trick ... just the plain FreeNAS installation ...
well ... not really what I want.

Anyone ANY idea at all?

 

[jgreco]

It isn't clear what the nature of the issue here is. I'd suggest experimenting a little and seeing if it's really "all jails" etc, or just some.

 

[Silvan Burch]

hi jgreco

it is all jails, doesn't matter which one I activate. but more concerning is that even just setting a static route in freenas results in the same ...
so it's probably network related ... is it possible that the hardware (Intel C236, Intel i210+Intel i219 LAN) could be the problem? the release notes from FreeNAS 9.10 say different but idk ...
it's a strange problem ... only thing related that I found was on the pfsense forum and pfsense is also based on FreeBSD ...

https://forum.pfsense.org/index.php?PHPSESSID=ddve1nin6gal13t04rdupsed90&topic=102216.15

 

[Silvan Burch]

I found the service that produces the problem:

watchdogd!

when I deactivate it manually, i.e. 'service watchdogd stop' the problem occurs sometimes ... but when disabled during runtime, no problems at shutdown!

Question is: what exactly does watchdog and is it save for me to disable it?

if someone knows the answer to this, I'm happy! :)

thanks for your help guys

 

[jgreco]

It's there to reboot the system if the system gets hung or stuck. In theory the watchdog firing is a bad thing because it suggests something else has gone awry. But if your system is being happy without it, that's kinda bad-wrong too.

I see you updated the bug report already, so let's see what devs say.

 

[Silvan Burch]

ok ... so it should actually prevent what it causes now ... strange.
anyway, as watchdog itself seems to be the problem (remember that the problem occurs during shutdown of the watchdog daemon) I'll stick to that solution until someone knows the reason ... I can't figure it out.

thanks