At this point, our self-hosted CI and code review tooling has had <i>significantly</i> better uptime for the past year.<p>The advantage of having someone else operate it evaporates when it is neutralized by complexity and scale. The failure modes of our own setup are much, much easier to deal with.
Do you think it makes sense to go all in on GitHub Actions without a backup plan involved? It's time to contact the CEO of GitHub again.<p>Don't say I didn't warn you for months and months on end. [0]<p>[0] <a href="https://news.ycombinator.com/item?id=26067426" rel="nofollow">https://news.ycombinator.com/item?id=26067426</a>
The “subscribe for updates” functionality is nice. Do other services offer things like this for updates on incidents?<p>That’s much better than watching Twitter or refreshing a page!
Is it just me or are these happening a lot more since Microsoft?<p>We unfortunately made the call to go all in in actions; and it really makes me miss gitlab...<p>The runners don’t even speak kubernetes properly; all your steps run inside the same container as the runner process...
They switched to<p>- Azure<p>- dotnet<p>Results:<p>- Lower uptime<p>- Degraded performance<p>Let's not forget about the global XBOX LIVE outage from last week
What's especially annoying is that most PRs seem to get a green checkmark for me as if checks have passed, when tests actually just haven't been started. It seems like it ought to be possible to do better than this. Right now the risk of accidentally merging a PR that hasn't run through CI is way too large.