Installation of shoot fails after about garden-setup HOT 2 CLOSED

Hm, there are multiple things here:

First, the error message you quoted doesn't come from garden-setup, but from the Gardener itself. So if the timeout waiting for the worker nodes is too short, this has to be fixed in Gardener or one of its extensions.
The only time garden-setup creates a shoot is when shooted seeds are configured. However, the shoot is created with Gardener and garden-setup has a timeout of 1h waiting for it to be finished - shoots usually come up within this timeframe if there is no error.

Regarding detection of previously created artifacts: If you don't touch the state folder, garden-setup should not lose track of kubernetes resources that it creates. It doesn't create resources on the infrastructure directly, apart from the etcd backup bucket, which it also keeps track of. Which artifacts did you have problems with?

from garden-setup.

I should report it in the gardener repo then.

Unfortunately, Its not so obvious which part of this huge stack of software is responsible for a failure - in particular for the timeouts.
The challenge is to identify all relevant states of the system in case of the failure

• we have the physical cloud cluster
• an OBS bucket for etcd
• the export, gen & state folders of the garden-setup.

If one is out of sync, only radical deletion would allow me to carry on.
I believe to have encountered leftover fragments in all those places at some point.

Most people using installers dont have a lot of detailed knowledge in the beginning.......

from garden-setup.