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
I’m automating the onboarding of accounts and other resources to Prisma Cloud using provider-http. As an initial step, I successfully created a role. However, when I tried to delete the managed resource (MR), it got stuck in the delete stage. After reviewing the pod logs, I noticed that a GET call is continuously triggered after the DELETE request. The external resource (the role) is deleted correctly, but the MR remains stuck. The GET/OBSERVE call is failing because the external resource was already deleted as part of the original DELETE request. I came across a similar issue reported here. I've tried several configurations in expectedResponseCheck, but none of them have worked so far.
Just to explain, the provider sends a GET request after a DELETE to confirm the resource was deleted. Right now, it only supports cases where this GET returns a 404 status. In your case, it looks like the GET is returning a 400, which is causing the issue. I understand the need for more flexibility here.
What happened?
I’m automating the onboarding of accounts and other resources to Prisma Cloud using provider-http. As an initial step, I successfully created a role. However, when I tried to delete the managed resource (MR), it got stuck in the delete stage. After reviewing the pod logs, I noticed that a GET call is continuously triggered after the DELETE request. The external resource (the role) is deleted correctly, but the MR remains stuck. The GET/OBSERVE call is failing because the external resource was already deleted as part of the original DELETE request. I came across a similar issue reported here. I've tried several configurations in expectedResponseCheck, but none of them have worked so far.
pod logs
describe output after delete request is made
How can we reproduce it?
Please find the manifest which used for role creation.
What environment did it happen in?
Crossplane version: v1.17.2
Include at least the version or commit of Crossplane you were running. Consider
also including your:
kubectl version
)uname -a
)Cc: @barunavo
The text was updated successfully, but these errors were encountered: