GitHub Actions Degraded Performance | Hacker News Reader

GitHub Actions Degraded Performance

githubstatus.com

26 pointsinfomiho8 comments

Open articleView on HN

Four days in a row at GitHub with different outages, someone's definitely having a difficult on-call.

- https://www.githubstatus.com/history

there have been, according to the github outage history page, more than thirty this year.

Microsoft's slowly making the case for hosting a gitea or gitlab on a dedicated server.

> Microsoft's slowly making the case for hosting a gitea or gitlab on a dedicated server.

I actually recently setup my own instances of Gitea (for source code), Nexus (artefact storage) and Drone (CI/CD): https://blog.kronis.dev/articles/goodbye-gitlab-hello-gitea-...

It wasn't too hard and although in my case I migrated over all of my projects from GitLab, the same should work with GitHub, or most others as well.

That said, while it is pretty cheap, lightweight and not too complicated, all of the sudden you're responsible for the configuration, uptime, updates and security.

Since I've basically nothing of too much value (and have a variety of backups) that's not an issue for me, but others might have different circumstances.

That said, if you like more integrated solutions and can afford a beefier VPS/VM then GitLab is also great, albeit the issues I had with updates on the Omnibus install urged me to move onwards.

Has there been any write-up or comment on why these happen so often? At this point Github outages are getting ridiculous.

We post monthly availability reports https://github.blog/tag/github-availability-report/

Whichever team of SRE is on rotation hopefully will get some extra cool-off time.

I had my workflow hitting github actions limits today that I'd never seen before, because the workflow was making many more api calls than usual. Did anyone else experience this?

Yep, got a notice to that effect for the first time but didn't confirm that it's related to the outages.

Reply on news.ycombinator.com