Make It Clear When a Deployment/Upgrade Fails in The KOTS Admin Dashboard

[jimmybrancaccio]

Is your feature request related to a problem? Please describe

About 30 minutes ago, I attempted to update my Gitpod instance from main.2860 to main.2898. I see the following:

If I hadn't also been running watch -n5 kubectl get po -n gitpod on my laptop, I wouldn't have known this had failed. It also does not properly reflect the actual version running on my cluster (which is still main.2860.

Describe the behaviour you'd like

I would like to know if the upgrade process failed along with information on why it failed and what needs to be done for it to successfully run.

Describe alternatives you've considered

N/A

Additional context

I was able to get a describe on the installer pod - https://ghostbin.linuxbox.ninja/paste/nr4wy

[corneliusludmann]

@mrsimonemms: Is this fixed with the following pull requests?

• [kots]: add an installation status pod #9455
• [kots]: allow deployment whilst installer job running #9471

[mrsimonemms]

@corneliusludmann yes, this is LARGELY achieved by these PRs. The only one that isn't is why it failed, but there isn't an easy way how to get that out of KOTS. I'm going to proposing closing it, but I'm going to add a ticket to run DB/Storage/Registry connectivity checks in the pre-flight checks (I know Replicated say that's not what they're designed for, but I think they're essential for how we're using it)