Big sack o' source GitHub is having a hellish week as the Microsoft tentacle suffered wobbles aplenty even as it tipped the scorn bucket over the emissions of the US administration.
Having fallen over in dramatic style on 21 April, seen its notifications totter on 22 April, and had trouble with Actions Workflows in the small hours of 23 April, the platform decided to take an extended lunch break today.
Twitter was its usual supportive self as developers found themselves faced with the dread error code 500 and a humorous depiction of the GitHub mascot tumbling into a ravine (like the unfortunate Wile E Coyote of Looney Tunes fame).
GitHub itself recognised that there were "issues" at 13:20 UTC. By 13:33 UTC, engineers reckoned they had found the source of the borkage and were hurriedly plugging the servers back in working on a fix.
Then again, we don't know too many developers who would refuse an extended luncheon (or extra morning coffee, depending on the timezone) so a short interruption may not be the end of the world. Access to Vulture Central, at least, seemed to return shortly before 14:00 UTC.
The issue looked to be global, although the timing meant that much of the US remained in blissful ignorance while Europe and the rest of the world wailed.
The issues come after GitHub made all core features of the platform available for everyone, declaring that "every developer on earth should have access to GitHub. Price shouldn't be a barrier."
Stability, on the other hand, does seem to have presented the house of code with a few issues recently, raising the unpleasant spectre of scalability. Certainly, new owner Microsoft has famously struggled to meet the recent unprecedented levels of demand on its own cloud.
As for the current topple?
We have deployed a fix and are continuing to closely monitor services as we observe recovery. https://t.co/WrxCrV1TEb— GitHub Status (@githubstatus) April 23, 2020
Until, of course, the next time. ®
* Total Inability To Service User Pulls