Downdetector's methodology is pretty garbage. I'm always surprised to see techies referencing them.
How is it garbage? It's crowdsourced monitoring so sure, there will be false positives (especially when aws/cloudflare/azure/bgp services have a bad day), but it's one of the best indicators to quickly know if _something_ is going on.
At the other extreme is Microsoft's own dashboard (https://azure.status.microsoft/en-us/status) whose methodology is so "rigorous" that it probably won't show anything about today's outage until next week, if at all!
It seems like one of the best methodologies there is? Real people reporting real outages of a service they use?
I often know about outages dozens of minutes in advance of it being posted on official channels (or in the case of AWS, three days later...).
So what if the resolution is low because people can't log in to Fortnite and an Azure outage gets shown as an Epic Games outage. Doesn't matter. The live feed is incredibly reassuring from a "down for anyone else this instant, before official status page has a chance to see it?" Saves me so much time knowing whether it's a me-or-them issue.
Still more reliable than most of the status pages that even lie after the outages
As someone who ends up being unofficial downdetector-primary at a company, dealing with this shit and the fact that people cannot comprehend its mis-attribtion wastes so much of my time.
not really, soon as i had a problem with outlook i went there and already saw user reports and a massive spike confirming it wasn't anything on my end.
It's something to reference. My view is that it's better than "doesn't load for me".