GitHub appears to be struggling with measly three nines availability

Devs say GitHub keeps glitching, blame Azure move and 'enshittification'

TLDR: GitHub hit a rough patch with outages, delays, and Copilot glitches, and users say uptime even fell near 90%, with a status-page revamp making trends hard to see. Commenters blame the Azure migration and Microsoft, share broken workflows, joke about “GitDown,” and debate self‑hosting as teams plan for outages.

Developers are roasting GitHub after yet another string of stumbles, turning “five nines” (near‑perfect uptime) into the meme of the week. On Feb 9, Actions, pull requests, notifications, and Copilot wobbled, with notification delays up to 50 minutes. Copilot issues lingered into Feb 10, and a reconstructed status feed suggests uptime even dipped below 90% in 2025. A recent status‑page redesign? Users say it hides the big picture. Cue the pitchforks.

The hottest thread pins blame on the ongoing move to Microsoft Azure, with pilif sighing, “A migration like this is a monumental undertaking… probably to not do it.” Another commenter dunked on leadership’s 2025 promise—“availability is job #1”—calling the results predictable. One user, Eikon, said they disabled workflows after repeated shutdown errors. The snark kept coming: sammy2255 quipped, “still running on a single MySQL machine,” while rvz dropped the hammer: “GitHub has completely enshittified,” arguing self‑hosting now feels more reliable.

Defenders are scarce, but pragmatists note GitHub’s 99.9% uptime guarantee applies mainly to enterprise customers, not everyone, and warn teams to plan for downtime. The vibe? GitHub’s brand took a hit, Azure’s catching strays, and devs are workshop‑renaming it GitDown.

Key Points

  • On February 9, GitHub experienced issues affecting Actions, pull requests, notifications, and Copilot.
  • GitHub reported notification delays (~50 minutes initially, ~30 minutes by 17:57 UTC) and confirmed normal operations by 19:29 UTC on February 9.
  • Copilot had policy propagation issues from 16:29 UTC Feb 9 to 09:57 UTC Feb 10, preventing some newly enabled models from appearing.
  • GitHub changed its status page design, making 90‑day availability visualization more difficult; an unofficial status feed suggests uptime dipped below 90% at one point in 2025.
  • GitHub’s SLA for Enterprise Cloud specifies 99.9% uptime, but this is not guaranteed for all users, underscoring the need to plan for downtime.

Hottest takes

"GitHub has completely enshittified" — rvz
"A migration like this is a monumental undertaking... probably to not do it." — pilif
"I wonder if they are still running on a single MySQL machine" — sammy2255
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.