server's foo.local name stops working after a few days

[ddrucker]

I have a TrueNAS-12.0-BETA system at home. It gets its IP from my router, via DHCP - let's call it foo.local.

For several days, all machines on my network can access it at foo.local. Then, suddenly that name stops working. I can still access it via IP - just the name stops working.

No other machines on my network that get a something.local name have this problem, and I did not have this problem under 11.3.

Restarting avahi-daemon fixes the problem for a few more days, but it always fails again after a few days.

Any ideas?

 

[seanm]

Did you check the avahi logs?

 

[ddrucker]

Ok, interesting:

daemon.log:Jul 26 04:00:07 grauman 1 2020-07-26T04:00:07.943748-04:00 grauman.local avahi-daemon 3120 - - Withdrawing address record for 192.168.1.3 on em0.
daemon.log:Jul 26 04:00:07 grauman 1 2020-07-26T04:00:07.943783-04:00 grauman.local avahi-daemon 3120 - - Leaving mDNS multicast group on interface em0.IPv4 with address 192.168.1.3.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211201-04:00 grauman.local avahi-daemon 3120 - - Interface em0.IPv4 no longer relevant for mDNS.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211518-04:00 grauman.local avahi-daemon 3120 - - Joining mDNS multicast group on interface em0.IPv4 with address 192.168.1.3.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211536-04:00 grauman.local avahi-daemon 3120 - - New relevant interface em0.IPv4 for mDNS.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211614-04:00 grauman.local avahi-daemon 3120 - - Registering new address record for 192.168.1.3 on em0.IPv4.
daemon.log:Jul 26 04:15:02 grauman 1 2020-07-26T04:15:02.025183-04:00 grauman.local avahi-daemon 3120 - - Withdrawing address record for 192.168.1.3 on em0.
daemon.log:Jul 26 04:15:02 grauman 1 2020-07-26T04:15:02.025295-04:00 grauman.local avahi-daemon 3120 - - Leaving mDNS multicast group on interface em0.IPv4 with address 192.168.1.3.


This happened this morning at 4am, which, almost certainly not coincidentally, is when my weekly jails backup happens, which involves iocage stop/export/start (no other relevant commands, everything else is just mv/cp).

And indeed, doing an iocage stop/start on my jails causes this problem to recur, until I do another restart of avahi-daemon.

So now what? Do I just add a restart of avahi-daemon to my backup script, or is there an underlying issue to fix here?

 

[anodos]

Ok, interesting:

daemon.log:Jul 26 04:00:07 grauman 1 2020-07-26T04:00:07.943748-04:00 grauman.local avahi-daemon 3120 - - Withdrawing address record for 192.168.1.3 on em0.
daemon.log:Jul 26 04:00:07 grauman 1 2020-07-26T04:00:07.943783-04:00 grauman.local avahi-daemon 3120 - - Leaving mDNS multicast group on interface em0.IPv4 with address 192.168.1.3.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211201-04:00 grauman.local avahi-daemon 3120 - - Interface em0.IPv4 no longer relevant for mDNS.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211518-04:00 grauman.local avahi-daemon 3120 - - Joining mDNS multicast group on interface em0.IPv4 with address 192.168.1.3.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211536-04:00 grauman.local avahi-daemon 3120 - - New relevant interface em0.IPv4 for mDNS.
daemon.log:Jul 26 04:00:08 grauman 1 2020-07-26T04:00:08.211614-04:00 grauman.local avahi-daemon 3120 - - Registering new address record for 192.168.1.3 on em0.IPv4.
daemon.log:Jul 26 04:15:02 grauman 1 2020-07-26T04:15:02.025183-04:00 grauman.local avahi-daemon 3120 - - Withdrawing address record for 192.168.1.3 on em0.
daemon.log:Jul 26 04:15:02 grauman 1 2020-07-26T04:15:02.025295-04:00 grauman.local avahi-daemon 3120 - - Leaving mDNS multicast group on interface em0.IPv4 with address 192.168.1.3.


This happened this morning at 4am, which, almost certainly not coincidentally, is when my weekly jails backup happens, which involves iocage stop/export/start (no other relevant commands, everything else is just mv/cp).

And indeed, doing an iocage stop/start on my jails causes this problem to recur, until I do another restart of avahi-daemon.

So now what? Do I just add a restart of avahi-daemon to my backup script, or is there an underlying issue to fix here?

Go ahead and file a bug report specifically mentioning that the start / stop of jails triggers the issue. We may do something relatively simple like have jails network changes restart avahi.

 

[ddrucker]

Done, https://jira.ixsystems.com/browse/NAS-106950