Support force delete default setting option on the account and project level
under review
V
Visiting Puffin
We have a scenario where people are using a process to deploy but it's not actually creating an instance in kube or anything like that but shows in Harness as having an existing deployment. They are unable to delete their service due to the existing instance that doesn't exist... We are trying to avoid giving them force delete option at the account level.
Log In
Prateek Mittal
Hi Cole,
What are you trying to deploy and is it something can be deleted on the infrastructure side? Is aborting of the pipeline not helping with this.
V
Visiting Puffin
Prateek Mittal - If I am recalling correctly, this issue was around Apigee deployments. We have teams use a process that runs an SSH script to deploy their application to Apigee, not Kubernetes. Harness is tracking that as a deployed service running on some kube infrastructure even though the SSH script was only ran on kube through the delegates (running on Kube). The same way any SSH scripts are ran in Harness.
So when their deploy completes there is nothing left on the kube side to clean up but the service still thinks there's a running instance and won't let you delete the service.
I don't believe there were any running deployments when we tried to delete the service. Just that the previous deploy and the instance tracker on the service showed something was deployed, even though it didn't exist on the kube side anymore.
V
Visiting Puffin
I will say I only saw this once and can't remember which team/project had the issue but it was a problem. We could probably close this until it comes up again because it sounds like it might just be an edge case.
Prateek Mittal
under review
V
Visiting Puffin
In addition to this we would also like the ability to control force delete resources at the project level. We do not want to give all users the ability to force delete references but rather give them the ability to turn it on/off themselves inside their project.