Unscheduled system reboots since upgrade to TrueNAS

[iXô]

Hi,

it took me a long time to do the upgrade, but finally I did it, I went from FreeNAS 11.2-U8 to TrueNAS 12.0-U3 a few days ago.

Before the upgrade, I had an uptime of more than 400 days, and now the system don't stay up more more 2 days.

I disabled the ipmi watchdog (stopped the watchdogd service and put a cron task to stop the watchdog) but it doesn't seems to be the problem, althought I am not so sure.

Last night I received this email :

Current alerts:
* tyco had an unscheduled system reboot.
The operating system successfully came back online at Fri Apr 30 03:38:46 2021.

If I search for /var/log/message at around this time, I see that :

Apr 30 00:00:00 tyco syslog-ng[1448]: Configuration reload request received, reloading configuration;
Apr 30 00:00:00 tyco syslog-ng[1448]: Configuration reload finished;
Apr 30 03:40:43 tyco syslog-ng[1441]: syslog-ng starting up; version='3.29.1'
Apr 30 03:40:43 tyco ---<<BOOT>>---

A ipmitool sel list return this :

1 | 04/30/2021 | 01:39:08 | Watchdog2 #0xca | Timer interrupt () | Asserted

The file /data/crash/info.last have that :

Dump header from device: /dev/da11p1
Architecture: amd64
Architecture Version: 4
Dump Length: 1219584
Blocksize: 512
Compression: none
Dumptime: Fri Apr 30 03:36:35 2021
Hostname: tyco
Magic: FreeBSD Text Dump
Version String: FreeBSD 12.2-RELEASE-p6 f2858df162b(HEAD) TRUENAS
Panic String: page fault
Dump Parity: 1396480542
Bounds: 2
Dump Status: good

I don't understand the last one, could this be the problem ? could this be a RAM problem ?


Here is my system (it's a little bit old now) :

SUPERMICRO X10SL7-F-O - CM µATX SINGLE SKT LGA1150
INTEL XEON E3-1231 V3 3.40GHZ LGA1150 BOX
CRUCIAL 16GB (8GBX2) DDR3L 1600MT/S DR ECC * 2

 

[ThreeDee]

running any jails?

 

[iXô]

running any jails?

Yes, only one jail, running Plex pass. I have updated it after the upgrade to TrueNAS, both the jail system, and Plex.

 

[ThreeDee]

for testing .. having tried stopping your Plex jail and see if you get any reboots? and what is TrueNAS installed on? SSD?

 

[iXô]

for testing .. having tried stopping your Plex jail and see if you get any reboots? and what is TrueNAS installed on? SSD?

Before mentioning the jail, I didn’t thought it could be it. I will see if there is a reboot this night and then try after stopping it.

The system is installed on a raidz pool consisting of 2 SD cards accessible via 2 SD to USB adapter, both on the same 4 Port usb hub plugged on the motherboard internal USB port. It works 24/7 since 2015.

 

[ThreeDee]

TrueNAS 12 accesses boot drive more so than older versions did ... hence recommendation of SSD's. I used to run a 120GB SSD via SATA to USB cable for quite awhile without issue ..

 

[iXô]

Thanks, I will see if the jail was the culprit and maybe switch the boot storage. Can I replace it like when I replace a failed hard drive, one at a time, even on a completely different bus ?

 

[ThreeDee]

Thanks, I will see if the jail was the culprit and maybe switch the boot storage. Can I replace it like when I replace a failed hard drive, one at a time, even on a completely different bus ?

I switched from my SSD connected with a SATA to USB cable (powered by the USB port only) to a 128GB NVMe M.2 drive by mirroring and then eliminating the SSD drive .. but you can just as easily back up config and then reinstall too