ci: run terraform apply in provider example test with timeout

[daniel-weisse]

Context

GitHub actions are run with a 6 hour internal timeout, after which running steps are terminated.
The termination signal is not properly forwarded to the running Terraform process, leading to resources not being cleaned up and state information potentially being lost.
For us, this only happens when there is an internal error during the constellation apply stage of the terraform provider, which is ran without any timeouts.
Since terraform apply does not support any global timeout flags, we may hit the 6 hour mark during this step.

Proposed change(s)

• Run terraform apply with timeout to stop execution of apply after 1 hour, or 4 hours for upgrades
• Ignore lock file when running terraform destroy

Related issue

• Should fix these clean up failures caused by hitting GitHub's timeout:
  • https://github.com/edgelesssys/constellation/actions/runs/11866687017/job/33073770054
  • https://github.com/edgelesssys/constellation/actions/runs/11866687017/job/33073769769

[netlify]

✅ Deploy Preview for constellation-docs canceled.

Name	Link
🔨 Latest commit	5cb6129
🔍 Latest deploy log	https://app.netlify.com/sites/constellation-docs/deploys/673d96d6e1d4550008c23f8a

[3u13r]

this only happens when there is an internal error during the constellation apply stage of the terraform provider, which is ran without any timeouts.

This would also affect other people using our provider in the Terraform CI, right? Moreover, I think the proper fix is to have a default, user editable timeout in the constellation apply step.
Having said that, we can do this as a follow up, since we want to fix CI issues here.

[3u13r]

LGTM