For about a month now I've been experiencing periodic freezing of unRAID where the web UI and any services become totally unresponsive or not reachable, the computer itself stays turned on though. Even SSH doesn't reach and my only course is to hard-reset the server by power-cycling it.

 

I can't think of any changes in my setup apart from re-seating the GPU and re-creating a VM after I've been experiencing a 127 error. This was ultimately fixed by updating to 6.11.1 a week ago. Since the update I've had the longest uptime of over a week but experienced a freeze again just now.

I have also removed a 50mV undervolt on my CPU at the same time thinking it might be related but this newest crash tells me it may not have been.

 

I've attached today's diagnostics and one older one from around the time the problem started happening. I've also enabled the syslog server (sadly, had it misconfigured before so it didn't catch the old crashes) and am waiting for the next crah(es).

 

---

 

This system was an upgrade from my last AMD 1900X system and it being 11th gen I expected there to be some issues but it's literally been bug-hunting for me every other month since my upgrade in January (latest find was an issue with correctly passing through the iGPU). The system:

• Intel 11600K (UHD 750 passed through to linuxserver/Jellyfin)
• 2x16 GB DDR4 ECC
• GTX 1650 (in W11 gaming VM, ACS override enabled)
• 3x 12TB in array
• 1TB cache SSD
• 4TB and 2TB unassigned devices (known bad sectors, just used for game storage and other unimportant media)

tower-diagnostics-20221013-1043.zip

tower-diagnostics-20221009-0314.zip

Edited October 14, 20223 yr by DesertCookie

Enable the syslog server and post that after a crash.

Finally got a crash again, @JorgeB. Since I am unsure in which log it'll be, I've uploaded both of the past few days.

syslog-1665646066 syslog-1666178969

Oct 17 04:45:37 Tower kernel: macvlan_broadcast+0x10a/0x150 [macvlan]
Oct 17 04:45:37 Tower kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]

Macvlan call traces are usually the result of having dockers with a custom IP address and will end up crashing the server, upgrading to v6.10 or later and switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)).

Thank you. I'll try this.

 

I use Docker containers with fixed IPs for containers like PiHole and Jellyfin that I want to prioritize in my router. Otherwise they get really slow, for instance, when Duplicacy is eating up bandwidth.

@JorgeB So far so good but I've noticed something changing from macvlan to ipvlan:

 

While before containers would appear as separate computers in my home network, allowing me, for example, to prioritize them in my router without having to prioritize the entire traffic to my server, these containers no longer appear. They do work as they are reachable under their assigned IP address; but my Fritz!Box doesn't pick them up as separate entities anymore, not allowing me to prioritize them, which was the whole reason I did forego the host and bridge network types.

Sorry, can't really help with that as I don't use custom networks with docker.