You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There's little assurance that the charts' renderings are coherent for a given set of input values. The current minikube test assures only that helm upgrade succeeds for a single set of input values.
This limited test doesn't protect against rending invalid manifests for a different set of input values. For example, the minikube test specifies ingress resources for the nginx controller, but another valid set of input values that does not use ingress at all may still render as an invalid configuration.
The CUE framework seems to have what we'd need to enforce a spec for a particular manifest, possibly a set of manifests, and a higher-level tool that leverages CUE is https://timoni.sh/ and may be worth looking at.
Any work we do in this area must be transferable to a future where an OpenZiti Operator is generating manifests, supplying input values to Helm, or both.
The text was updated successfully, but these errors were encountered:
qrkourier
changed the title
test charts' renderings
test charts' renderings for coherence per type of deployment
Mar 19, 2024
There's little assurance that the charts' renderings are coherent for a given set of input values. The current minikube test assures only that
helm upgrade
succeeds for a single set of input values.This limited test doesn't protect against rending invalid manifests for a different set of input values. For example, the minikube test specifies ingress resources for the nginx controller, but another valid set of input values that does not use ingress at all may still render as an invalid configuration.
The CUE framework seems to have what we'd need to enforce a spec for a particular manifest, possibly a set of manifests, and a higher-level tool that leverages CUE is https://timoni.sh/ and may be worth looking at.
I found Timoni mentioned in Reddit: https://www.reddit.com/r/kubernetes/comments/1be15b8/comment/kuu6bco/?utm_source=share&utm_medium=web2x&context=3 in the context of a discussion about the disadvantages of templating YAML manifests with Helm.
Any work we do in this area must be transferable to a future where an OpenZiti Operator is generating manifests, supplying input values to Helm, or both.
The text was updated successfully, but these errors were encountered: