Okay, I am spiking out a POC to migrate our traefik default ingress controller to HA using consul as a backend. I have consul up and running in a test namespace and am at the point where I want to use the storeconfig command to put our traefik.toml values into the kv store the easy way... However, as soon as the pod starts it dies with this error: Error: failed to start container "traefik-consul-test": Error response from daemon: OCI runtime create failed: container_linux.go:345: starting container process caused "exec: \"storeconfig\": executable file not found in $PATH": unknown
I am using the :1.7.12-alpine image and below is the relevant specs of my deployment object, I've left out the volumes, env vars, and other sensitive info:
and got this back still: Error: failed to start container "traefik-consul-test": Error response from daemon: OCI runtime create failed: container_linux.go:345: starting container process caused "exec: \"traefik storeconfig --consul --consul.endpoint=consul-test-consul-server-0:8500\": executable file not found in $PATH": unknown
Funny thing is that I can kubectl run -i --tty traefik image=traefik:1.7.12-alpine -- sh and run the command from within the shell just fine, which I suppose satisfies my needs, but doesn’t answer the question.
Okay, you’re still my fave, but I am still hitting the same error… Error: failed to start container "traefik-consul-test": Error response from daemon: OCI runtime create failed: container_linux.go:345: starting container process caused "exec: \"storeconfig\": executable file not found in $PATH": unknown
^ that was using traefik:1.7.12 without the traefik command. Tried it just as you have configured above on the alpine image… Same error.
To be clear, it’s having a hard time connecting to consul still - but at least we have the k8s deployment manifest commands working. I can handle the networking side on my own.
And this one works for using the traefik:1.7.12 image: