rd-michel
rd-michel
just fyi "Terraform v1.0.0 intentionally has no significant changes compared to Terraform v0.15.5"
we're also looking forward to use variables in lifecycle blocks. example: set prevent_destroy to true per default ... overwrite it in our dev environment on a case by case basis
confirmed. ive tested this setup ubuntu 2204-lts + k8s 1.26.1 works (fixes also cgroup v2 problem - ubuntu 2204 vs k8s 1.21.0)
i think a proper diff view is really neccessary here. e.g. kube-prometheus-stack helm chart contains a approx 3000 line values file. changing one line of code outputs 6000 lines of...
nothing changed for me with tf 1.8.0 example: * deploy kube-prometheus-stack helm chart * afterwards add a single new line to the helm values file * rollout again result: ❯...
@towolf we always use the latest version of the helm provider (2.13.1 atm). why is it necessary to downgrade to 2.9.0?
Thanks @towolf Ive pretty much the same problem as described by @tanadeau The displayed helm values context for large helm values files can be overwhelming during a terraform plan/apply Perfect...
current workflow: 1. rollout terraform code and find out that i complains about an existing "prevent_destroy" 2. raise a PR to set prevent_destroy=false for e.g. all existing gcp clusters just...
+1 terraform-ls using 700% cpu while working on a big repository with multiple terraform projects/directories.
for sidecar patterns this is still an important feature. feedback appreciated. maybe your recommendation is to migrate away from oauth2-proxy sidecar pattern setups