I'm aware of the usual way to "tail --follow" logs of a Kubernetes service:

kubectl logs -f -lapp=service-name --all-containers=true

But every time I push a change to this service the kubectl dies when the pods get killed and I have to run it again.

Example error message:

Received interrupt, cleaning up... rpc error: code = Unknown desc = Error: No such container: b48a1ad3c0080680465c79903d03748a026becf397bc780921674b2f0d7078ffReceived interrupt, cleaning up...

rpc error: code = Unknown desc = Error: No such container: 6258e31702ea678eacec5ad0df15b5620b5609cd5e4822f2e4991fd26c9906b6 $

What I wonder is if there is a way to tell Kubernetes to keep looking for new pods which match the -lapp=service-name tag and tail them. A bit similar to tail --follow=name --retry

I suppose I can run the command in a simple shell loop but was wondering if there is something smarter that can use Kubernetes, one which will avoid lots of errors while the new pods get deployed.