I feel like failed resources should never stop a

pulumi refresh

from running. Is there an issue filed for that?

https://github.com/pulumi/pulumi/issues/2352 ?

@incalculable-sundown-82514 yeah, we had a resource deleted by hand so now I have to edit the state manually

gotcha! re: editing the state, can you use this command? https://pulumi.io/reference/cli/pulumi_state_delete.html

yeah, its just id prefer the CI/CD do it rather than by hand

gotcha.

thanks for the follow up

yeah,

pulumi state delete

is only useful for one-time disaster recovery. no problem!

yeah, we had a resource deleted by hand so now I have to edit the state manually

What kind of resource?

refresh

should be able to handle deleted resources.

Diagnostics:
  kubernetes:extensions:Ingress (external-rpc):
    error: Preview failed: 2 errors occurred:

    * Resource 'external-rpc-2hxv283a' was created but failed to initialize
    * Ingress .status.loadBalancer field was not updated with a hostname/IP address.
        for more information about this error, see <https://pulumi.io/xdv72s>

error: an error occurred while advancing the preview
make: *** [refresh] Error 255

One of the underlying target groups was deleted so the Ingress resource never fully built the ALB/ELB out

Ah, got it. This is exactly https://github.com/pulumi/pulumi/issues/2352.