ajchiarello

I see this same issue with several charts (such as bitnami/harbor), and it is not reproducible when using helm directly - I can upgrade the chart fine when using helm....

@itsecforu - It would be great to know what you had to set in your nginx settings.

I think that this issue can be tackled via the new [replace_triggered_by](https://devdosvid.blog/2022/05/04/new-lifecycle-options-and-refactoring-capabilities-in-terraform-1-1-and-1-2/#trigger-resource-replacement-with-replace_triggered_by) lifecycle argument presented in Terraform 1.2 - I haven't tested that yet, but it looks promising.

Here's the BIG-IP resources. In this, I do have a port defined under the virtual_server, because one is required to create the resource; as I said in the original post,...

Is there any timeline for resolving this bug?

@trinaths How would one go about following up on the bug ID?

Honestly, the delay on this led me to move from this operator over the [CloudNative PostgreSQL Operator](https://github.com/cloudnative-pg/cloudnative-pg), which does everything and is already secure.

This problem has just started occurring for me, with rancher 2.5.2. Changing context to address the kubernetes control plane directly does not experience the hang.

> I circumvented the problem with a MutatingWebhookConfiguration and the corresponding controller: I circumvented the problem by moving to the [CNPG Operator](https://cloudnative-pg.io/) which does everything securely by default.