Website returning a 522 error from Cloudflare, servers for JOSM uploads not responding, looks like I will be waiting for a while…
Wiki’s working fine, though.
Website returning a 522 error from Cloudflare, servers for JOSM uploads not responding, looks like I will be waiting for a while…
Wiki’s working fine, though.
Bei mir ist es auch so - auch ich kann OSM nicht laden, auch nicht via JOSM
“It’s the same for me - I can’t load OSM either, not even via JOSM”
Yep, so it’s not just me. I wonder if the techs are hard at work on this right now…
It’s down since at least 2 hours.
Yes, since approx. 2:30 UTC: Grafana
It’s back! … at least the website. API still times out for me.
Hm…habe die Website für einen kiurzen Moment gesehen - jetzt geht aber wieder nichts.
And I need the API up because I have a pending edit in JOSM right now.
This (522 errors) has been going on for me in California for several hours (since about 3:00 UTC). It’s possible to browse map data, but problems come when one tries to “open for writing” to OSM’s API, like a new iD session or a JOSM save.
In Grafana it looks like something nasty might have been going on.
Web requests were increased from 01:25 to 01:45.
Until 01:15 only oauth2 gets requested. And then from 01:45 until 02:30 suddenly other authentication types than oauth2 get requested, even though they have been turned off I think:
At 02:30 the Apache connections and threads spike, and then logging stops. (This is for spike-06
but looks similar for 07 and 08.)
Thanks for posting those nice graphs and your analysis. Looks like something rather unusual and serious (deliberate?) happened especially in the final 50 minutes or so, between 0145 - 0235.
Possible DDoS attack?
The “weird spiky traffic patterns” do lean in me agreeing with that direction of initial speculation, but I’m what might be called “only semi-professional” in this sort of analysis.
It seems to be working for some time now.