ever since updating too 7.0 im randomly loosing connection to my server, as if someone is ripping the ethernet cable out.. im having a hard time pinpointing what is happening i do know that when it happens my internet is up and stable. i can also see the DHCP lease, and i am able too ping my unraid machine via another local machine. im hoping someone here can help me pinpoint what is causing this. when this happens unraid connect doesnt catch it and still reports the machine as UP but wont let me manage. the only fix i have found as of now is restting my router states table or a manual reboot. here is a diag. thanks in advance

 

poseidon-diagnostics-20250212-1814.zip

Edited February 13, 20251 yr by Lassley7
updated diag because the system disconnected once again.

hmm i do see this happening alot in the sys log. is this normal activity?

Poseidon kernel: br-2b0f514dcdc9: port 10(veth91f1544) entered disabled state
Feb 12 16:16:05 Poseidon kernel: veth923cc83: renamed from eth0
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth91f1544) entered disabled state
Feb 12 16:16:06 Poseidon kernel: veth91f1544 (unregistering): left allmulticast mode
Feb 12 16:16:06 Poseidon kernel: veth91f1544 (unregistering): left promiscuous mode
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth91f1544) entered disabled state
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth996ffa0) entered blocking state
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth996ffa0) entered disabled state
Feb 12 16:16:06 Poseidon kernel: veth996ffa0: entered allmulticast mode
Feb 12 16:16:06 Poseidon kernel: veth996ffa0: entered promiscuous mode
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth996ffa0) entered blocking state
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth996ffa0) entered forwarding state
Feb 12 16:16:06 Poseidon kernel: br-2b0f514dcdc9: port 10(veth996ffa0) entered disabled state

 

Edited February 12, 20251 yr by Lassley7

It is when containers are starting or restarting, it's not normal if it goes on 24/7, enable the syslog server and post that after a crash.

4 hours ago, JorgeB said:

It is when containers are starting or restarting, it's not normal if it goes on 24/7, enable the syslog server and post that after a crash.

I will do this. But I am assuming that everything thing else in the diag looks ok? 

Nothing else jumps out in the syslog, assuming it covers an event, since it starts over after every boot.

3 minutes ago, JorgeB said:

Nothing else jumps out in the syslog, assuming it covers an event, since it starts over after every boot.

Sorry I should have explained that the above diag was when the system hung up and I didn’t reboot only thing I did was reset the router states to get back in. If nothing else jumps out I know this is probably unrelated but can a container issue cause this? My plex containers has been on the ritz due to a corrupted db but I don’t see that causing the system to hang up? 

Do you remember the aproximate time you did the router reset?

25 minutes ago, JorgeB said:

Do you remember the aproximate time you did the router reset?

Sometime between 7am-10am yesterday morning. Sorry I can’t be more precise 

Probably around this time:

 

Feb 12 09:25:37 Poseidon nginx: 2025/02/12 09:25:37 [error] 11327#11327: OCSP responder timed out (110: Connection timed out) while requesting certificate status, responder: r11.o.lencr.org, peer: 23.64.114.214:80, certificate: "/boot/config/ssl/certs/certificate_bundle.pem"

 

But still don't see anything relevant logged, if resetting the router resolved the issue, the server may not be the problem.

15 minutes ago, JorgeB said:

Probably around this time:

 

Feb 12 09:25:37 Poseidon nginx: 2025/02/12 09:25:37 [error] 11327#11327: OCSP responder timed out (110: Connection timed out) while requesting certificate status, responder: r11.o.lencr.org, peer: 23.64.114.214:80, certificate: "/boot/config/ssl/certs/certificate_bundle.pem"

 

But still don't see anything relevant logged, if resetting the router resolved the issue, the server may not be the problem.

Idk if this is in the diag logs but I am seeing this error as well? Could this be causing it? More importantly what is this error? P.s for some reason I can’t wrap the code via my phone.. 

```oseidon nginx: 2025/02/12 09:25:37 [error] 11327#11327: OCSP responder timed out (110: Connection timed out) while requesting certificate status, responder: r11.o.lencr.org, peer: 23.64.114.214:80, certificate: "/boot/config/ssl/certs/certificate_bundle.pem" Feb 12 10:30:25 Poseidon kernel: WARNING: CPU: 10 PID: 1254906 at fs/notify/fdinfo.c:55 show_mark_fhandle+0x77/0xe6 Feb 12 10:30:25 Poseidon kernel: CPU: 10 PID: 1254906 Comm: lsof Tainted: P O 6.6.68-Unraid #1 Feb 12 10:30:25 Poseidon kernel: Call Trace: Feb 12 Feb 12 16:04:45 Poseidon kernel: PMS BTQ[1786413]: segfault at 14c500000068 ip 000014c5deeb6d0f sp 000014c5addb0558 error 4 in ld-musl-x86_64.so.1[14c5dee6f000+53000] likely on CPU 10 (core 4, socket 0)```

Edited February 13, 20251 yr by Lassley7

That error means the server was unable to connect with the internet.

 

The call trace is irrelevant, all users get it once per boot, it's a kernel issue.

15 minutes ago, JorgeB said:

That error means the server was unable to connect with the internet.

 

The call trace is irrelevant, all users get it once per boot, it's a kernel issue.

So something is kicking the server off/on from the internet. Does the diag show my network settings? Do they look correct? I did jump quite a few versions. So. I wonder if I need to update my network settings 

They look fine to me, and if it was a network config issue, I would expect the server to always have problems, not just after a few hours, looks more like an external problem to me.