Server becomes unreachable every few days

[Abstract7227]

swapped my motherboard for a ASRock X570M Pro4 a few days back, already got a crash however.

CPU is next, PSU after that

 

I was wondering if a faulty OS flash drive could also throw an error like that? 

[Abstract7227]

So I did everything. Replaced RAM, CPU, motherboard, PSU, even installed new power cables to make sure they are not the cause. Even used a different power outlet.

 

The system is still crashing, found one warning in the logs (attached), after googling I found this issue

 

 

so if I understand this correctly, using the Custom : br0 with a static IP address causes issues?

Currently, all of my docker containers have a custom static IP address assigned to them via br0.

Am I to understand I have to bridge them all?

 

I don't quite understand why though, is this a bug in unraid?

T issue.txt

[JorgeB]

13 hours ago, Abstract7227 said:

br0 with a static IP address causes issues?

Switching to ipvlan should fix that (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right))

[Abstract7227]

I tried that, rebooted my server, got another crash today (logs attached) with btrfs corruption warning.

At this point I:

 

replaced my RAM (twice)

replaced my motherboard

replaced my CPU (twice)

replaced my PSU

replaced my cache drives

 

The only original parts now are the boot flash drive, the spinning drives (which are xfs formated) and the case.

Assuming I can rule a hardware issue out at this point, I am not sure how I should proceed now.

 

I only run a nextcloud instance with a swag reverse proxy and a home assistant instance on there, so it's not like I am doing a lot of crazy things with this.

 

Any recommendations?

 

 

issue.txt

[JorgeB]

Problem is with the filesystem on this device:

device dm-4

Can't tell which disk/pool that is based on the posted snippet.

 

[Abstract7227]

so I have to replace that one? cause that drive is pretty new

 

[Abstract7227]

This might be a noob question, but why is it that a btrfs error result in my server not being accessible anymore via the webgui (500 error)?

[JorgeB]

The snippet you've posted only shows filesystem issue, not device problems.

[Abstract7227]

and that can crash the entire server?

[JorgeB]

Not usually.

[Abstract7227]

cause I think this is the 3rd time that the whole server crashes while only spitting out a btrfs error. Anything else I can investigate?

 

[JorgeB]

Post current diags mostly to see the hardware used.

[Abstract7227]

My current diagnostics (after reboot, as I cannot get my diagnostics during the crash)

trantor-diagnostics-20230309-1712.zip

[JorgeB]

You are running the server with overclocked RAM, that's a known issue with Ryzen, and something I've already posted on post #2 on this thread.

[Abstract7227]

Ah damn you're right, I apologize.

 

I probably missed it because I went from 2 rank, 2/4 3200Mhz, which was fine, to 2 rank, 4/4 3200Mhz. I configured it to 2666Mhz now.

 

Thx for the feedback

[Wastl]

On 2/19/2023 at 9:41 AM, Abstract7227 said:

hrow an error like that? 

 

On 3/4/2023 at 11:09 PM, Abstract7227 said:

So I did everything. Replaced RAM, CPU, motherboard, PSU, even installed new power cables to make sure they are not the cause. Even used a different power outlet.

 

The system is still crashing, found one warning in the logs (attached), after googling I found this issue

 

 

so if I understand this correctly, using the Custom : br0 with a static IP address causes issues?

Currently, all of my docker containers have a custom static IP address assigned to them via br0.

Am I to understand I have to bridge them all?

 

I don't quite understand why though, is this a bug in unraid?

T issue.txt 6.62 kB · 1 download

You could try to remove the Unraid br0 network in the GUI and instead create a docker macvlan through the commandline or portainer.