Crash of knot-resolver with 203

After updating an ipfire box from 202 to 203 the box refused to do DNS resolving.

Searching for the reason it turns out that the knot-resolver did crash, error log is attached.

Further investigation regarding the crash results in line 130 of /usr/lib/knot-resolver/config.lua, this line is called from a funktion where /var/ipfire/dhcp/setting is read.
After disabling the following setting from
ENABLE_BLUE=on
to
ENABLE_BLUE=off
and restart knot-resolver DNS did work again as expected.

I don’t know where it comes from that in dhcp settings the Blue network was enabled, because the box runs only with Green and Red interface.

Additional information: On another ipfire box with 203 the setting ENABLE_BLUE=on doesn’t let crash knot-resolver. All other settings in /var/ipfire/dhcp/settings between the two boxes are the same, so I have actually no idea why the crash happens on the first ipfire box.

I would appreciate any ideas to shade a light on this issue.

Here is the log of the crash:

Jul 22 10:15:11 ipfire-tk kresd[9026]: [system] path = /run/knot-resolver/control/0
Jul 22 10:15:11 ipfire-tk kresd[9026]: [cache ] Cache top initialized using existing data (generic).
Jul 22 10:15:11 ipfire-tk kresd[9026]: [defer ] Initializing defer…
Jul 22 10:15:11 ipfire-tk kresd[9026]: [defer ] Defer initialized (generic).
Jul 22 10:15:11 ipfire-tk kresd[9026]: [defer ] Defer configuration: Expected cpus/procs: 4 Max waiting requests: 64.0 MiB Request timeout: 1.0 s Idle: 1.0 ms UDP phase: 400.0 us Non-UDP phase: 400.0 us Priority levels: 25 (4 main levels, 8 sublevels) + UDP KRU capacity: 524.3 k (4.2 MiB) Decay: 0.012 % per ms (32-bit max: 524280) Half-life: 5.7 s Priority rise in: 22.7 s Counter reset in: 90.8 s Rate limits for crossing priority levels as single CPU utilization: 1 2 3 max v4/32 : 0.012 % 0.195 % 3.125 % 49.999 % v4/24 : 0.391 % 6.250 % 100.000 % 1599.976 % v4/20 : 3.125 % 50.000 % 800.000 % 12799.805 % v4/18 : 9.375 % 150.000 % 2400.000 % 38399.414 % v6/128: 0.012 % 0.195 % 3.125 % 49.999 % v6/64 : 0.024 % 0.391 % 6.250 % 99.998 % v6/56 : 0.037 % 0.586 % 9.375 % 149.998 % v6/48 : 0.049 % 0.781 % 12.500 % 199.997 % v6/32 : 0.781 % 12.500 % 200.000 % 3199.951 % Instant limits for crossing priority levels as CPU time: 1 2 3 max v4/32 : 1.0 ms 16.0 ms 256.0 ms 4.1 s v4/24 : 32.0 ms 512.0 ms 8.2 s 131.1 s v4/20 : 256.0 ms 4.1 s 65.5 s 1048.6 s v4/18 : 768.0 ms 12.3 s 196.6 s 3145.7 s v6/128: 1.0 ms 16.0 ms 256.0 ms 4.1 s v6/64 : 2.0 ms 32.0 ms 512.0 ms 8.2 s v6/56 : 3.0 ms 48.0 ms 768.0 ms 12.3 s v6/48 : 4.0 ms 64.0 ms

Jul 22 10:15:11 ipfire-tk kresd[9026]: [system] error while loading config: /usr/lib/knot-resolver/config.lua:130: attempt to index local ‘address’ (a nil value) (workdir ‘/run/knot-resolver’)
Jul 22 10:15:11 ipfire-tk supervisord: exited: kresd0 (exit status 1; not expected)
Jul 22 10:15:12 ipfire-tk supervisord: gave up: kresd0 entered FATAL state, too many start retries too quickly