Discussion around spegel

[farazkhawaja]

First off — spegel is an excellent project and solves the embedded registry use case nicely(important usecase). However its a single-maintainer project that powers a critical path in every k3s cluster.

We've seen a few regressions lately in spegel that are breaking and causing blockade to cluster upgrades to latest k3s versions.

Has the team considered any alternatives for solving the embedded registry use case? Dragonfly and Kraken exist but are far too heavy for what k3s needs. Spegel really is the best fit today — which is partly why we're asking about the long-term plan around it.

Would love to hear the team's thoughts. Thanks!

[brandond]

We like it too, but do not at the moment have cycles to drive it's maintenance, or ensure that it's goals align with all the use cases that the k3s community may envision. I believe @phillebaba also only has limited time to review issues and PRs, as his day job and work on the "enterprise" version of Spegel allows.

Both are open source projects, so feel free to open PRs to address whatever is important to you. Maintainers on both side will do our best to respond.

[Starefossen]

May I suggest contributing to the spegel project? As you said it is clearly the best option - let's show some love and support to the maintainer ❤️

[phillebaba]

Hey @farazkhawaja, thank you for bringing awareness to an issue I also have been trying to bring attention too!
Spegel is a single maintainer project and with that, in true open source nature, I'm trying to rally the community to have more maintainers. On top of that, anyone with an enterprise use case that relies on Spegel for their uptime and business, I encourage them to advocate within their company to support Spegel financially by either Github donations or purchasing the enterprise version which includes SLAs.

In the meantime, I will keep working on making Spegel usable for the home lab and individual contributor use case. 🙂