Upgrade from TrueNAS core to Scale resulted random reboots

[shakezilla]

I recently upgraded from TrueNAS core to the most recent Scale release (22.02) following this guide and everything seemed to work fine until I got to the idle screen the system first rebooted. It rebooted a few more times after hitting the idle screen. Then it stayed on for a half hour or so the next time so I looked around and the logs didn't appear to be interesting at all (no signs of the crash before the boot messages began). While the system is running, however, my datapool is intact and accessible via command line, but my SMB shares are unfortunately unavailable (though this is likely due to an issue with the permissions).

I tried to delete the configuration in scale to see if it was a broken config import. Unfortunately, my system still crashes. Since deleting my config, I have not yet imported my dataset because I see no point while the system is as unstable as it is. In the meantime I've booted back into Truenas CORE 12.0-U8 because I don't want to keep putting the extra stress on my components.

Hopefully someone can help, this is my hardware:
Intel Xeon E-1220 V3 3.1GHz w/ Stock Fan
ASRock Z97 Extreme6 ATX LGA1150 Motherboard
2 x G.Skill Ripjaws X Series 8GB DDR3-1866 CL9 RAM
WD Black 250GB M.2-2280 NVME SSD
Silicon Power A55 256GB SSD
6 X WD Red 10TB 5400RPM (WD100EFAX) HDDs
EVGA GeForce 210 1GB DDR3 Video Card
Intel EXPI9301CT PCIe x1 1 Gbit/s Network Adapter
FSP Group 400 W 80+ Gold Certified ATX Power Supply
Fractal Design Define R5 ATX Mid Tower Case

My current configuration is 1 boot pool with the two SSDs mirrored, booting from the NVME. The other pool is a Raidz2 pool with all 6 of the 10 TB drives.

Each time the system reboots I have a simple alert that says "Unscheduled System Reboot". I'm not really sure where to go from here but I would really like to transition to a linux environment if possible.

 

[shakezilla]

Small Update

Since I reverted back to my previous configuration with CORE my system actually had another reboot while transcoding something in plex. This is the first time I've ever had a random or unexpected reboot with CORE that I can recall and I've been running this system for several years now. There's still nothing of note in the logs prior to the boot messages (just like the reboots in scale) and I really have no idea what's going wrong or where else to check. I assumed it was a software issue because the issue never presented itself in CORE before today AND it started up within minutes of my upgrade to scale. But now that I've reverted back to CORE and saw another very similar reboot I'm beginning to think it's hardware related after all. I have no idea why the issue would only begin to present itself recently but nothing else is really making any sense.

I'm getting desperate for ideas now because I need at least one solid system - up until now that was my CORE install. But I'm certainly not an expert in Linux or FreeBSD so I'm quickly running out of things to try.

 

[shakezilla]

This morning around 5AM my server rebooted again, unexpectedly. This time I did not have any of my jails running so I really have no idea what's going on. Interestingly, this time I did not get an alert in my log about "Unscheduled System Reboot". In fact, the only way I could tell my system had rebooted last night is that the uptime this morning reported as just a few hours. I checked the logs and again there's nothing of note prior to the system reboot messages.

Maybe I should just order a new mobo/cpu/ram/psu and hope for the best. Any thoughts on that plan? Or how I could maybe narrow it down a bit?

 

[shakezilla]

Update: It seems like the PSU must be going bad. I've been staying in TrueNAS core to try to keep my system as stable as possible and noticed that I was able to reliably trigger a reboot by playing something in plex that requires transcoding. I did a little research and decided to download mprime and run a torture test. My system died and rebooted within one second of starting the test.

Anyway, I just replaced the PSU with a EVGA Supernova 750W P5 Platinum and I've rebooted the system and it's been running for a few hours now. I've run mprime again and let it go for ~30 minutes. Everything remained stable. I've been playing shows off my plex since then to see if that triggers anything and so far everything has been fine.

If the system remains stable for the next few days I'll try again to boot into my TrueNAS SCALE boot sector. But I'm really hoping that my issues will be resolved with this hardware swap.

 

[artlessknave]

you might want to check out the power supply resource. i would rate your 400W PSU as insufficient to begin with.
likely it was barely running and age reduced its output to not enough. it's an FSP, so it probably lasted far longer than a less quality one would have.
your new 750W should not have that problem.

 

[shakezilla]

you might want to check out the power supply resource. i would rate your 400W PSU as insufficient to begin with.
likely it was barely running and age reduced its output to not enough. it's an FSP, so it probably lasted far longer than a less quality one would have.
your new 750W should not have that problem.

I think you are correct. My 400W PSU was probably sufficient when I first completed my build back in 2016. But over the years I've swapped out the drives for much larger drives, replaced the USB boot device with an SSD/NVME combo, and I've added the network card. Somewhere along the line I think I pushed the limits of the PSU and essentially strained it to death over the following months/years. It still works at low loads but I don't know enough about the inner workings of a PSU to try to repair it so I'm going to be throwing it out now that the new PSU seems to be stable.

Can you link me to the power supply resource? I also assume my new 750W PSU will have plenty of power for my components but I am curious to see what it says.

 

[artlessknave]

its in my signature.

throwing it out now that the new PSU

I wouldn't automatically throw it out, it would still probalby work fine as long as you are below 350W max load or so, so you could use it for a smaller draw load, like, a basic test bench or kinds PC or something.

definitely do not try to repair it. some of the caps hold very large charges, and shorting the wrong thing can lead to a significant zap. also, you do NOT want to plug in one of these that isn't 100% percent wired correctly.