Opam repository CI: being migrated and updated

Dear all,

The CI server that powers the opam-repository CI is being upgraded to a more powerful 36-core machine with fast disks. As part of this migration, the deployment will also be run by the bactrian bot, instead of my personal mugshot on the CI page :slight_smile:

I’m also taking this opportunity to fix several small ails in the opam2 integration, thanks to patches from @jpdeplaix. In the short-term, you may see duplicated status updates in the opam-repository PRs, since both systems are running simultaneously. I’ll post here when the migration is debugged and complete.

4 Likes

Great news! Also…

…I presume SSDs? (Pure curiosity.)

The new server is setup, but the migration to a proper bot user is currently blocked due to GitHub API rate limits. So both servers are still running concurrently. If anyone knows someone at GitHub that can help us expedite the transition by upping rate limits please let me know – I’ve mailed their support channel in the meanwhile.

1.6TB, NVMe, Mixed Use Express Flash – quite a satisfactory setup for I/O :wink:

Very nice indeed! I imagine it should be possible to keep the CPUs nearly fully occupied with that as the storage back end. I’ll be curious as to whether that turns out to be true in practice or not.