Kernel panic and boot problems

[dvand]

Hello all,

I tried to reboot my freenas server today. When booting up it gets to the point of zfs pool import and then I get a kernel panic.

I'm running the latest 9.10.2-(u6). Machine has been running flawlessly for years.

I'm at a loss as to what to do next. I am currently running memtest86 just in case there's memory issue.

Any ideas?

 

[dvand]

I'm not sure what is going on. After trying many dozens of times it went past the pool import and then the machine powered itself off.

I tried booting again and I got another kernel panic. Image attached.

 

[Chris Moore]

I'm at a loss as to what to do next. I am currently running memtest86 just in case there's memory issue.

Any ideas?

It looks like a hardware fault, but we need to know more about your hardware to be able to give good advice. Please reference this post to see what information we typically need to see:

https://forums.freenas.org/index.php?threads/updated-forum-rules-8-3-16.45124/

 

[dvand]

Here's what i have:

supermicro x10sl7-f
16GB of the approved ecc ram (2x8 sticks)
6 4TB drives in z2 configuration
an intel i3 (can't recall model number)

Any advice appreciated.

 

[Chris Moore]

I'm not sure what is going on. After trying many dozens of times it went past the pool import and then the machine powered itself off.

Before we do any further testing, disconnect your data drives so we don't have any potential of causing damage there.
Once you have disconnected the drives, try booting the system to see if FreeNAS will boot without the drives. It is possible that a defective drive might cause the system to fail. When I have seen that in the past, it prevented the system from completing POST, but it is worth testing.
Also, can you update your profile to at least show what country you are in. It helps if I know whether or not you can run down to Best Buy and grab a new hard drive or cable.

 

[dvand]

disconnected the drives and attempted to boot.
Get a different crash.

 

[Chris Moore]

Machine has been running flawlessly for years.

What is the boot device?

 

[Chris Moore]

I'm at a loss as to what to do next. I am currently running memtest86 just in case there's memory issue.

Did you get any results from the MemTest?

 

[dvand]

Unfortunately not, Memtest didn't show any errors.

 

[dvand]

Unfortunately not, Memtest didn't show any errors.

I'm starting to think motherboard is flaky? Just tried a boot again and it booted right up (without drives installed).

Not sure what to try next?

 

[Chris Moore]

Not sure what to try next?

I have seen some strange behavior when the boot device starts to go bad and you said it had been running for years. That is why I asked what the boot device was. If it is up for the moment, backup your system configuration so you can do a fresh install to a new boot device and restore the config DB. Do you know the steps? Here is the section of the manual:
http://doc.freenas.org/9.10/system.html#advanced

 

[dvand]

Sorry missed the boot device question. I have two USB drives in mirror. SanDisk cruzer fits.

I tried flashing to a fresh USB drive earlier today and the drive faulted out.



Sent from my Pixel XL using Tapatalk

 

[Chris Moore]

Sorry missed the boot device question. I have two USB drives in mirror. SanDisk cruzer fits.

I tried flashing to a fresh USB drive earlier today and the drive faulted out.

So, you removed both of the original USB drives and you were not able to install to a fresh USB drive because of some fault.
As a test, if you have a small hard drive that you are not using, can you try installing any operating system to the NAS. It is sounding like a failed system board, but I hate to make that call too quickly. I feel like more testing would be good.

I have extra computers all over my house because, reasons, but I know that isn't true of everyone. Do you have a spare computer laying about, that you can test the original FreeNAS boot media with. I just want to see if it will boot reliably.

 

[dvand]

I can try to boot it off a laptop.

That being said, i just booted again and it went past importing the pools and bringing systems up, then a whole bunch of stuff whizzed by on the console (looked like errors or faults) and then the system powered off by itself.

That seems pretty odd.

 

[Chris Moore]

then a whole bunch of stuff whizzed by on the console (looked like errors or faults) and then the system powered off by itself.

That seems pretty odd.

That is one of the things you said, "powered off by itself", that makes me think it is a hardware fault in the system board. What do you have connected at this point?
What power supply are you using. I have also seen strange behavior caused by bad power from an aging supply

 

[dvand]

At this point i have the 6 hard drives connected. It's a seasonic power supply.

Looks like it's hanging at post now with message "System Initializing" with code 15 displayed.

I'm thinking I need to buy another board.

 

[Chris Moore]

It's a seasonic power supply.

More details please. It could still be a bad power supply and I hate to have you buy a system board if it is actually the power supply causing the problem.
Do you have any other desktop computer power supply that you can test with?

 

[dvand]

Seasonic ssr-440rt active pfc f3 if I'm reading it right.

Sent from my Pixel XL using Tapatalk

 

[Chris Moore]

Machine has been running flawlessly for years.

Seasonic ssr-440rt active pfc f3 if I'm reading it right.

If you have another power supply, you might want to try changing that as a test. If it is still misbehaving, I would certainly look hard at the system board. The other thing you might try is just using one memory module at a time, but you said that passed the MemTest.
I have only ever seen a few processors fail and all of those were from overclocking / not being properly cooled. If you have a good cooler (do you?) and you are not overclocking, I can't see the CPU having failed.
I just hate to send you straight to replacing the system board. Is it possible that it might still be under warranty?

 

[dvand]

The CPU has never been overclocked nor ever run hot when i've checked it. I have a (really) old computer i might be able to use for a power supply to rule that out.