[v2] improve the log messages

[postmaxin]

Seeing a lot of stuff like this;

{"level":"error","ts":1603747129.250157,"logger":"controller","msg":"Reconciler error","controller":"ingress","name":"rex-ingress-external","namespace":"","error":"cannot find at least two subnets from different Availability Zones, discovered subnetIDs: []"}
{"level":"error","ts":1603747155.3905587,"logger":"controller","msg":"Reconciler error","controller":"ingress","name":"rex-ingress-dev","namespace":"","error":"Service \"qualia-sync-cron-service\" not found"}

[M00nF1sh]

@postmaxin
It's expected. I assume you are using rex-ingress-external and rex-ingress-dev as IngressGroups. these are encoded as a types.NamespacedName without namespace.

Do you need help on the actual error?

1. the first one should be either u didn't tag the subnets correctly or u didn't set the '--cluster-name` flag to controller correctly
2. the second one is the service referenced by one Ingress in IngressGroup rex-ingress-dev is not found

[postmaxin]

So for #1, that totally makes sense, and I'm fine.

For #2 -- it was looking for the service in a namespace, correct? How do we tell what namespace it was looking in?

[M00nF1sh]

I agree we should improve the logs for #2. Before that, you can check all ingresses with group.name rex-ingress-dev :D.

[postmaxin]

Oh boy! :D

$ kubectl get ing -A | grep rex-ingress-dev | wc -l
   158

Ok, I'll do that for now

[M00nF1sh]

@postmaxin
We'll track the fix in v2.0.1

[weberc2-tempus]

the first one should be either u didn't tag the subnets correctly or u didn't set the '--cluster-name` flag to controller correctly

How can we debug the --cluster-name flag setting?

[M00nF1sh]

@weberc2-tempus the --cluster-name setting is a flag to the controller, we currently didn't print it out in the logs. maybe in future versions, we can echo them.