The DNS Sidecar fails when one of the container instances is not running

[cptanalatriste]

✅ Checklist

• I have searched open and closed issues for duplicates.
• This is a problem observed when using a Data Safe Haven.
• I can reproduce this with the latest version.
• I have read through the documentation.
• This isn't an open-ended question (open a discussion if it is).

💻 System information

• Operating System: macOS
• Data Safe Haven version: v5.7.1
• Browser details: Safari

🚫 Describe the problem

For multiple reasons, like saving costs, sometime TRE admins manually stop some of the TRE container instances (i.e. Gitea, HedgeDoc). When they do that, this breaks the DNS Sidecar functionality:

This prevents DNS repair for the rest of the containers.

🚂 Workarounds or solutions

For a permanent solution, we need to implement container status checks in the DNS Sidecar script. As a workaround, you can manually start all the stopped containers, to then manually trigger the DNS Sidecar.