I'm at an end here.

My unraid install has been crashing / becoming unresponsive / losing connectivity more frequently as time goes on.

 

So far, I've done the following:

• Replaced my Zigbee stick to a Sonoff ZBdongle-P
• Changed my docker from MACVLAN to IPVLAN
• Replaced my boot drive with a Kingston DTSE9H
• Upgraded my RAM to Corsair Vengeance LPX 32GB (2x16GB)
• Slotted my boot drive in a USB2.0 slot

 

Most of the things I've done (aside from upgrading the Zigbee stick) is due to posts I've read about boot drives failing, hardware failure and docker issues with MACVLAN in newer releases.

 

None of the above has solved the issue.

I'm running a bunch of docker containers for media and home automation.

The home automation part is especially painful.

 

I've attached the current syslog, because the server is once again in this weird state.

 

These kinds of messages start to pop up, and after that, Unraid & all dockers lose internet connectivity. Local management is sometimes still accesible, sometimes not.

 

Jan  4 10:13:46 Tower kernel: traps: dotnet[1624] general protection fault ip:15531cab9602 sp:7ffe28a4d6f0 error:0 in libc-2.31.so[15531cab9000+15a000]
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth32e3052) entered disabled state
Jan  4 10:13:46 Tower kernel: veth4a518a5: renamed from eth0
Jan  4 10:13:46 Tower  avahi-daemon[5233]: Interface veth32e3052.IPv6 no longer relevant for mDNS.
Jan  4 10:13:46 Tower  avahi-daemon[5233]: Leaving mDNS multicast group on interface veth32e3052.IPv6 with address fe80::8896:7eff:fe9e:728a.
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth32e3052) entered disabled state
Jan  4 10:13:46 Tower kernel: device veth32e3052 left promiscuous mode
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth32e3052) entered disabled state
Jan  4 10:13:46 Tower  avahi-daemon[5233]: Withdrawing address record for fe80::8896:7eff:fe9e:728a on veth32e3052.
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth0c07beb) entered blocking state
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth0c07beb) entered disabled state
Jan  4 10:13:46 Tower kernel: device veth0c07beb entered promiscuous mode
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth0c07beb) entered blocking state
Jan  4 10:13:46 Tower kernel: docker0: port 1(veth0c07beb) entered forwarding state
Jan  4 10:13:46 Tower kernel: eth0: renamed from veth739ecb9
Jan  4 10:13:46 Tower kernel: IPv6: ADDRCONF(NETDEV_CHANGE): veth0c07beb: link becomes ready
Jan  4 10:13:48 Tower  avahi-daemon[5233]: Joining mDNS multicast group on interface veth0c07beb.IPv6 with address fe80::78b5:83ff:fe9a:ae61.
Jan  4 10:13:48 Tower  avahi-daemon[5233]: New relevant interface veth0c07beb.IPv6 for mDNS.
Jan  4 10:13:48 Tower  avahi-daemon[5233]: Registering new address record for fe80::78b5:83ff:fe9a:ae61 on veth0c07beb.*.

 

EDIT: Attached diagnostics aswel.

 

tower-syslog-20230104-0914.zip

tower-diagnostics-20230104-1025.zip

Edited January 4, 20233 yr by Crisicus

Those messages by themselves are not necessarily a problem, but you can shutdown all docker containers and enable one by one to see if you find which one is doing it.