dashboard
dashboard copied to clipboard
Non-user friendly message when attempting to delete Cloud Credentials being used
Setup
- Rancher version: 2.7-head (Commit ID: 4092002)
- Browser type & version: Brave: v1.44.108 (Chromium: v106.0.5249.103) and Firefox v105.0.3
Describe the bug When attempting to delete a Cloud Credential that is being used by an active cluster, the user is presented with a non-user friendly error message.
To Reproduce
- Create AWS Cloud Credentials in Rancher
- Create an EC2 Node Driver cluster using the newly created AWS Cloud Credentials
- When the cluster is Active, attempt to delete the AWS Cloud Credentials
Result Note the user is presented with a non-user friendly error message in the "Are you sure?" dialogue box.
Expected Result A clear and concise error message is presented that explains to the user why they are not able to delete that set of Cloud Credentials.
Screenshots

When following the steps to reproduce on v2.7-head (Commit ID: 76c98b4), I was able to verify that the message has now been changed to the message: "Index with name by-cloud-cred does not exist". However, the message provided does not make sense in the context of the users actions (attempting to deleting a set of cloud credentials that are being used by an existing cluster). Therefore I am reopening this issue.

@jameson-mcghee There's nothing more the UI can do here as the message comes from the backend.
I am going to move back to test - in order to get the error message improved, can you please file an issue against the backend team. Thanls.
jameson-mcghee said: Based on the information provided by @nwmac, I am flagging this ticket as QA Blocked until the completion/validation of Rancher ticket 39300.
It was determined that since the remaining work related to this is issue will be done on the backend Rancher ticket 39300, this ticket can be marked as Done.
@jordojordo This fix doesn't seem to be present on 2.6-head, is there a way we can get it back ported? I'm unsure on how the UI team handles making fixes across both 2.6 and 2.7, but since this is a bug we'd like to resolve it for both versions. The backend portion has already been back ported.
@HarrisonWAffel We can backport it manually. I created an issue to track this, although we haven't branched for the next 2.6 release yet.