Kubernetes updating deployment image requires Ingress to be remade

[AlecTroemel]

Traefik version: v1.1.0

Currently, when the docker image of a kubernetes deployment is updated, the traefik route becomes broken and a new ingress needs to be created. I get these errors on updating a deployment as well

level=error msg="Undefined backend 'agentchat.mirus.io/example' for frontend my.domain.com/example" 
level=error msg="Skipping frontend my.domain.com/example..."  

Doesn't make sense that traefik ends up routing to a single pod (which is terminated and replaced on updating), since the ingress exposes a service, but thats my current thought from my understanding of the web ui

Any thoughts/solutions on this?

[emilevauge]

@errm any opinion on this ?

[AlecTroemel]

Updating to v1.1.1 resolved this issue, closing it!