Since Microshit's code is closed source, I am forced to assume they AI-ed the shit out of their shit, until they start practicing source code transparency.
post
When I write the software I'm working on I always add my bugs in plain sight - that makes them easier to fix. Hiding it was such a noob play, Microsoft!
If Lemmy wasn't so fucking rabidly anti-Microsoft and anti-AI, we might have an interesting discussion about this. But out of all the comments here, it's all hate.
I'm all for skepticism, but jesus christ everything is the fault of AI and everything is AI and it just absolutely ruins it all.
If the cause is as they said, this is just one of those things that happens. They were testing for problems, and testing on weaker hardware is a Good Thing™ because that's where a lot of problems happen. So this was an unusual bug. And if what they said is true about some of the mitigation - like extending email delivery times to cause fewer failed emails - they put thought into how to mitigate the damage.
But we apparently can't discuss any of those things. Everyone was just quick to jump on the MICROSOFT EVIL bandwagon and "don't put all your eggs in one basket" while that is exactly what you do when you choose your email provider, for example, no matter who you choose. Even with self-hosting, you're putting all your eggs into the basket of what you use to host it.
I don't want people to worship any particular tech. I personally like Windows....... and Linux........ and Andoird...... and while I don't care about anything Apple, for those that use those products and like them, enjoy! I generally don't feel a need to constantly shit on the things I don't personally like and prefer.
So yet again, with all the delightful things that happen on Lemmy, here's another topic that can't be discussed impartially because of the rabid unhealthy obsession most lemmings have on hating these things to the point that it completely ruins any sort of conversation on the topic.
Now I've ranted before, and I expect the downvotes, and I don't fucking care. I'm sure there will be replies about how this upsets me and I suck or whatever, and you can make those replies to make yourself happy, but I will not be engaging with them. My point, other than just venting my frustration, is in the tiny hope that maybe some might re-evaluate and take a fresh look at it and maybe in future we might have more reasonable conversations. Again, I'm not saying nobody should hate Microsoft, but jesus christ people, the over-the-top "they're completely lying" and blaming AI is just dumb.
Anyway. Whatever.
To save the curious among you some scrolling:
the change ran globally for more than three months without visible problems. [...] Microsoft says neither the defect nor that fix applies to general Windows deployments, separating this incident from a Windows desktop issue.
Inside an internal Windows process, competing tasks were waiting for access to shared resources, a problem called lock contention. Under certain traffic conditions, this exhausted processor and memory capacity.
Connection requests piled up until servers stopped responding. Microsoft’s tests missed the issue because they focused on weaker machines, while the fault became worse on newer hardware with more cores.
"Whew boy... I'm glad Microsoft was taken down by a 'hidden bug', rather than one of those run-of-mill bugs I'm sure they've got." /s
The "micro" in Microsoft is for their competency level.
And for the type of females Bill Gates is attracted to, just like his late good friend Epstein
Just FYI, try using "women" instead of "females" as the former is more inclusive. The latter is used more by people that are bigots and fascists.
Downvotes for calling out Bill Gates for being caught up in the Epstein files? Lots of Destiny-style liberalism finding its way onto .world these days I see
According to this they are funneling all traffic through a single "gateway". What the fuck?
No, it definately says "gateway servers" at least once.
The "Moving from the secondary gateway to the primary gateway" probably means "from the secondary gateway deployment to the primary gateway deployment".
They might have been A/B or blue/green deploying a new gateway system. Or perhaps the old gateway system was on outdated hardware and were transitioning to new gateway hardware (which lines up with "this impacted more powerful hosts more" kinda thing they say).
Even if it's a single virtual node per region, it would be fully load balanced across probably thousands of servers per region.
Its preliminary incident report, dated November 15
I want to read this, but I only see links to other articles on this site, and a couple of useless forum threads.
Interesting. A client side bug causes server-level overload but only after unrelated changes gradually make the problem worse. Rerouting traffic to healthy servers overloads those too and spreads the outage. The client code surely “worked on my computer” lol
Where did they say it was a client side bug?
I guess I misunderstood it.
Inside an internal Windows process, competing tasks were waiting for access to shared resources, a problem called lock contention. Under certain traffic conditions, this exhausted processor and memory capacity.
I read this as the behavior that precipitated the issue on the gateway was caused by how Windows sent traffic to exchange servers. So by Windows they mean whatever is running on the gateway?
It says multiple times that the issue was with Microsoft's internal cloud server software, not Windows.
The defect came from a fix prepared specifically for Microsoft 365 networking gateway servers.
The failure was within Microsoft’s cloud infrastructure, not customer Windows systems
Why do I find myself doubting this lol
Nobody is willing/allowed to admit the truth: "We used AI and it blew up in our face instantly" goes for the flights grounding, Iranian school, etc
It's a good question. Did you perhaps want it to be some other explanation, and so now that you got an explanation you didn't want you prefer to doubt it?
all 47 comments