Helm chart

[olevski]

Hi it would be really nice if there was a Helm chart for this.

I would like to contribute one and open a PR. Would you be willing to review and merge this? Also would you be willing to publish the image and helm chart on this github repo? You can publish the helm chart and have a helm chart repo with github pages.

[funkypenguin]

I too, would be interested in a helm chart, and I'm happy to help contribute towards this.

[olevski]

I was just reminded of this post by a colleague. I wanted to mention that we forked this repo (https://github.com/SwissDataScienceCenter/csi-rclone) at the place where I work and added a helm chart. Note that our fork has diverged and we have changed things quite a bit. We are currently implementing the ability to recover/remount disks when the pods of the daemon set restart.

Anyhow if any of the maintainers here are interested in merging / re-combining things let me know.

[funkypenguin]

Not a maintainer, but I'd love to know how the remounting/recovering is done :)

[olevski]

So the tldr is you save the rclone configurations for every mount in text file(s). Then when you restart you check and remount. There is a draft PR here: SwissDataScienceCenter#72. But this is still a WIP so the exact implementation will change from what is there but the core idea is the same.

[olevski]

One thing to note is that you will still interrupt users who may be using those mounts but at least the mounts will be back up shortly after a restart. Currently after a restarts all the mounts are just broken.