I feel like by now GitHub has a worse downtime record than my self hosted services on my single server where I frequently experiment, stop services or reboot.
I got Claude to make me the exact same graph a few weeks ago! I had hypothesized that we'd see a sharp drop off, instead what I found (as this project also shows) is a rather messy average trend of outages that has been going on for some time.
The graph being all nice before the Microsoft acquisition is a fun narrative, until you realize that some products (like actions, announced on October 16th, 2018) didn't exist and therefore had no outages. Easy to correct for by setting up start dates, but not done here. For the rest that did exist (API requests, Git ops, pages, etc) I figured they could just as easily be explained with GitHub improving their observability.
I mean I'm as annoyed as the next person about the outages but I'm not sure correlating with the Microsoft acquisition tells the whole story? GitHub usage has been growing massively I'd imagine?
I'm convinced one of my org's repos is just haunted now. It doesn't matter what the status page says. I'll get a unicorn about twice a day. Once you have 8000 commits, 15k issues, and two competing project boards, things seem to get pretty bad. Fresh repos run crazy fast by comparison.
It’s biaised to show this without the dates at which features were introduced. A lot of the downtimes in the breakdown are GitHub Actions, which launched in August 2019; so yeah what a surprise there was no Actions downtime before because Actions didn’t exist.
GitHub is 100x the size today with 100x the product surface area. Pre-Microsoft GitHub was just a git host. Now, whether GitHub should have become what it is today is a fair question but to say “GitHub” is less stable today vs. 10 years ago ignores the significant changes. Also, much of these incidents are limited to products that are unreliable by nature, e.g: CoPilot depends on OpenAI and OpenAI has outages. The entire LLM API industry expects some requests to fail.
GitHub’s reliability could stand to be improved but without narrowing down to products these sort of comparisons are meaningless.
When I say that Microsoft writes very bad code some people get offended. For example for Azure Event Hubs they have almost no documentation and Java libraries that mostly do not run.
I will chime in that Jira and Bitbucket have drastically improved performance and reliability over this same time period. It actually feels snappy and they seem to listen to feedback.
45 comments
[ 2.6 ms ] story [ 49.8 ms ] threadThe data is there, you just have to hover over each data point.
Looking at this now, you might as well self host and you would still get better uptime than GitHub.
[0] https://news.ycombinator.com/item?id=22867803
"The Missing GitHub Status Page" with overall aggregate percentages. Currently at 90.84% over the last 90 days. It was at 90.00% a couple days ago.
The graph being all nice before the Microsoft acquisition is a fun narrative, until you realize that some products (like actions, announced on October 16th, 2018) didn't exist and therefore had no outages. Easy to correct for by setting up start dates, but not done here. For the rest that did exist (API requests, Git ops, pages, etc) I figured they could just as easily be explained with GitHub improving their observability.
Maybe that's just the date when they started tracking uptime using this sytem?
GitHub’s reliability could stand to be improved but without narrowing down to products these sort of comparisons are meaningless.