Good day people, 

I was running on 6.11 (can't remember which version exactly) I recently decided to change my server chassis from Rosewill RSV-R4000U to RSV-L4412U, for swapping disks if ever needed since I was nearing 2 years of current HDD use, then I thought, new chassis, why not upgrade to Unraid 6.12.3 ? 

And so I did some research, tried to look for anything that I have to be aware of, but in the sea of information it was hard to find anything indicative of any failures and everything I read kept indicating that upgrading to 6.12 was a breeze and there will be no issues. 

and so I did, and now I regret, first issue was unraid didn't want to boot at all after the upgrade, I kept getting hardware error messages ( that I still get until now - attached below ) at boot, it was very hard to capture, because the moment that screen comes up, it would go blank about one second later and the system would just freeze, I had to long press the power button to get it back, but that happened more than once and I wasn't comfortable doing that so I decided to keep it powered down, and I just removed all disks then went online to try to find out the cause but it was really hard to pin point the issue.

I got a new USB dongle, I restored a backup from unraid connect, and still the same then I read somewhere online that unraid now can run using UEFI so I did try that and viola it doesn't throw a lot of errors and when it gets stuck the screen doesn't go blank or lose signal, then I decided to start in safe mode, and thankfully it did end up booting in safe mode and I can see my settings are intact, I read online to disable plugins and so I did and installed 4 plugins ( Unassigned Devices, CA, Dynamix System Temperature and Unraid Connect) , and then I restarted the server and it booted back normally and I started the array and all was well, until it froze some few hours later.

 

So I decided to do a fresh install and then move the config files back to the new install maybe that would change the outcome or so I thought. I continued reading online anyways  and I found something about macvlan vs ipvlan, I tried to find such an option on unraid on how to disable it but couldn't find it, so I just stopped using that one container that used a vlan setting, (And after this post I will remove vlan altogether from the Interface in network settings within unraid), rebooted and the server stayed up for 3 days all good until today when it also decided to freeze.

I have no clue why it does freeze but here is what I already did, I enabled syslog mirroring and I will attach it hopefully something useful would come out of it, I have no clue how to read the syslog, can't understand it even though I tried, I only saw one pattern which doesn't make sense but it kept saying error and throwing the PID to CPU 3, not sure if this is something related or not. 
I also attached diagnostics of after the crash.

even though I can ping the server, I cannot enter the GUI, the monitor has no signal so I will restart the server now, I will run a memtest.

The hardware is the same, I only added an HBA (DELL H310 6Gbps SAS HBA FW:P20 9211-8i IT Mode ) to be able to actually use more HDDs. 
Also now I am using a PiKVM V3, so I switch from the KVM Screen HDMI to the PiKVM when I leave home for a long period ( like when I did 3 days ago and only came today to find the server frozen although I did use a VM for cloud gaming around 4 hours before the crash )

At the moment the PC specs are as follows

Intel i7 11700

32GB DDR4 Ram (Corsair LPX)
Asus H570 Pro Motherboard
RTX 2060 Super (passed through for VMs) with a fake HDMI dongle attached
3 Seagate ironwolf nas hdds as parity and 2 data disks 8TB each
1tb nvme drive passed through for a VM
1tb SSD share
240gb ssd share

850W platinum PSU
 

looking forward for help. and thanks in advance.

syslog unraid-diagnostics-20230827-0047.zip

See the release notes, specifically this part:

https://docs.unraid.net/unraid-os/release-notes/6.12.0#crashes-related-to-i915-driver

9 hours ago, JorgeB said:

See the release notes, specifically this part:

https://docs.unraid.net/unraid-os/release-notes/6.12.0#crashes-related-to-i915-driver

Thanks will do and see what happens, may I ask where or how do you know it is an issue related to i915 Driver? I want to learn more about diagnosing / troubleshooting on my own. 

Ill report back if any crash happens after doing what is mentioned in the release notes.

7 minutes ago, BinAdhed said:

may I ask where or how do you know it is an issue related to i915 Driver?

The call trace mentions the iGPU driver:

 

Aug 26 16:08:25 UNRAID kernel: DPLL 0 assertion failure (expected on, current off)
Aug 26 16:08:25 UNRAID kernel: WARNING: CPU: 9 PID: 13092 at drivers/gpu/drm/i915/display/intel_dpll_mgr.c:191 assert_shared_dpll+0xbe/0xed [i915]

 

1 hour ago, JorgeB said:

The call trace mentions the iGPU driver:

 

Aug 26 16:08:25 UNRAID kernel: DPLL 0 assertion failure (expected on, current off)
Aug 26 16:08:25 UNRAID kernel: WARNING: CPU: 9 PID: 13092 at drivers/gpu/drm/i915/display/intel_dpll_mgr.c:191 assert_shared_dpll+0xbe/0xed [i915]

 

Thank you for that. 

I ran the command and will report back if it does crash again. 

happy to report that the system been up and running with no issues until now.

 

thanks

On 8/27/2023 at 1:16 PM, JorgeB said:

The call trace mentions the iGPU driver:

 

 

Unfortunately, the system decided to freeze today while I was doing a parity sync.

All I did was, stop array, shutdown, insert new HDDs, start server, replace old parity drive with new, start array.

It was going good, and last time I checked it said 13 hours remaining, only to find out later that the server is not responding again.

What should I do in this case?

I think log is still enabled on flash drive, should I attach the file?

1 hour ago, BinAdhed said:

should I attach the file?

Do it, may be something there.

2 hours ago, JorgeB said:

Do it, may be something there.

syslog attached, 

I have an assumption and might be wrong but do I need to retype the command from the release notes after every restart? is that command not permanent? 

syslog

Sep 12 22:31:51 UNRAID kernel: macvlan_broadcast+0x10a/0x150 [macvlan]
Sep 12 22:31:51 UNRAID kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]

 

You are having macvlan call traces, change docker network to ipvlan and reboot.

 

20 minutes ago, BinAdhed said:

I have an assumption and might be wrong but do I need to retype the command from the release notes after every restart?

Which command? If it's the one about the iGPU it's one-time.

 

3 hours ago, JorgeB said:

Sep 12 22:31:51 UNRAID kernel: macvlan_broadcast+0x10a/0x150 [macvlan]
Sep 12 22:31:51 UNRAID kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]

 

You are having macvlan call traces, change docker network to ipvlan and reboot.

 

Which command? If it's the one about the iGPU it's one-time.

 

would the macvlan still be an issue even if I don't use vlan at all?

thanks for the answer, yeah i meant the iGPU. so my assumption was wrong. 

I will change to ipvlan now and hopefully report back in a week or so of it going with no hiccups.

Thanks Jorge you helped a lot.

12 minutes ago, BinAdhed said:

would the macvlan still be an issue even if I don't use vlan at all?

Yes, if you get call traces they can end up crashing the server.