Hi @DevelopmentCats, could you please have a quick look when you have a chance?

@brymut, can you fix the failing CI? You may need to add an exception to the typos config.

https://github.com/coder/registry/actions/runs/16600024993/job/46963017795?pr=254#step:7:41

@matifali Failing CI is because of a variable being mistaken for a spelling mistake. Please check again

@brymut you need to add an exception to this in .github/typos.toml

@brymut you need to add an exception to this in .github/typos.toml

Sorry, I missed that exception handling, I've added the word to the list. Please re-run CI and have a look @matifali

I re-ran the CI but its still detecting it as a typo though. You might have to try setting the typo as just hel

Thanks for the feedback, I've tested locally and changing the typo exception to "hel" works. @matifali / @DevelopmentCats could you please re-run the CI

Left a few suggestions but looks good to me. I will approve after another review.

Applied requested changes @matifali

Hi @bpmct could you please have a look at this?

Maybe @hugodutka could have a quick look?

@brymut Hello, the team is pretty busy lately and we haven't been able to give proper attention to your PR.

That said, please refrain from pinging people (and especially at that rate), as it will not make it faster.
@bpmct and @hugodutka are not the go-to people for module reviews, and both are currently at capacity.
Atif (@matifali) has been busy lately, and will look into it when possible.

As you know, we are still a small company and things like this are tough to manage, but sending a lot of messages will only increase how hard it is for us to manage our workload. We are not ignoring your module intentionally, but do understand that with our different roles, there are things we have to prioritize.

I understand that this is important for you, but we are not able to review this instantly, and I will let @matifali and @DevelopmentCats handle this.

Thank you.

I'll give this a final review today. Sorry things have been crazy busy 🙂

I loaded this up in my template and tested it a few times and it seems to work well.

The only thing that I notice is that the parameters for the Server Location, and the Server Type's should be Dynamic Parameters so that you can make it only give the server types that are available for each location since the workspace will fail if you select a server type that is not available in a specific location without any meaningful feedback.

Here are the docs for dynamic parameters: https://coder.com/docs/admin/templates/extending-templates/dynamic-parameters

Let me know if you have any questions @brymut

I do also notice that there are some issues with the lifecycle of the VM. When the workspace is stopped the server in hetzner stays running and the workspace hangs after that since it is supposed to be stopping the hetzner vm, but it never does.

Also the labels were not being applied

Something else that I am noticing while testing this module is that you can only provision a single workspace with this template since there are not unique names for the volume or the server and when its used again it would error out because they exist already.

@brymut any update on this?

Hi, just pushed changes @DevelopmentCats. Dynamic params added, Hetzner cleans up as expected when workspace is shut & multiple workspaces can be started using the template. Issue regarding the cloud init label not applying to the home volume isn't solved yet but please have a look while I figure that out. Will update today

@DevelopmentCats label working on os fs level and added some labels at the hetzner console level too. I think this should all be ready to go now.

@DevelopmentCats label working on os fs level and added some labels at the hetzner console level too. I think this should all be ready to go now.

Sorry about the delay on this. Thank you for figuring out the label stuff. I will go ahead and test this again with the latest changes to make sure everything is working fine tonight.

Sorry about the delay on this. Thank you for figuring out the label stuff. I will go ahead and test this again with the latest changes to make sure everything is working fine tonight.

Created new recording with requested fixes @DevelopmentCats. This version shows the provisioning of multiple hetzner instances in coder with dynamic param enabled, running simultaneously and shown in the Hetzner console, checking labels on both the OS Filesystem and on the hetzner console and then shutting down through coder demonstrating clean up on hetzner and deleting workspaces without errors:

Sorry about the delay on this. Thank you for figuring out the label stuff. I will go ahead and test this again with the latest changes to make sure everything is working fine tonight.

Created new recording with requested fixes @DevelopmentCats. This version shows the provisioning of multiple hetzner instances in coder with dynamic param enabled, running simultaneously and shown in the Hetzner console, checking labels on both the OS Filesystem and on the hetzner console and then shutting down through coder demonstrating clean up on hetzner and deleting workspaces without errors:

hetzner-fix-1.mp4

Thank you so so much for providing the screen capture after updating the behavior. This is looking awesome overall, but the last question I have for this template in regards to hetzner is this. I think it might be worthwhile to maybe create a module for handling snapshot's with hetzner later on since as of right now none of the data will persist passed the workspace stopping from what I can see, but this can be done in another PR later on.

@brymut I removed count from the volume resource, since it was destroying the volume when the workspace stops.

This now has an optimal lifecycle:
Workspace Starts -> Volume Created, Server Created, Volume attached, Server Starts
Workspace Stops -> Server Stops, Volume Unattached, Server Destroyed
Workspace Starts Again -> Server Created, Volume Attached, Server Starts

Sounds good, if you can create a separate issue to track the module you described, I'd happy to take it on once this is merged. Bounty or not @DevelopmentCats

Anything else I can do here @DevelopmentCats ?