This message was deleted.
s
This message was deleted.
b
is your cluster unreachable?
m
Yes, it’s deleted
b
set the environment variables
PULUMI_K8S_DELETE_UNREACHABLE
and then run
pulumi refresh
1
PULUMI_K8S_DELETE_UNREACHABLE=true
m
Get the same error
I believe the state is very out of sync now. It doesn’t include resources that are there, and it includes resources that have been deleted/are unreachable. Are there any mechanisms for being able to clear this and run pulumi up? I’d be willing to manually go in and delete all the infra by hand, but want to ensure when
pulumi up
executes that it wouldn’t make false assumptions about certain resources being up with a non-empty state.
b
You must be on an old provider version, you'll have to manually remove the resources in which case
m
Right, I’ll do that on the infra side. But how do I clear the pulumi state for the stack? After I delete the infra, if I run
pulumi up
, it’ll still assume the k8s resources are there since they’re in the state.
b
you'll need to
pulumi stack export
modify the JSON to remove the resources then
pulumi stack import --file
or if you have the urns, you can do
pulumi state delete <urn>
m
export --> import probably would have done the trick. Didn’t let me delete urns since it claimed they didn’t exist. Ended up deleting the stack and redeploying.
513 Views