Update: Everything* appears to be working so far, thanks @jerry
*Except for images, see this post. Fedia never got images working on day one. It would probably explain our problems.
Update 2: I made this comment a while back. I don’t like how it’s aged… details
It’s been over half a day since the last post from a non-fedia user came through, and while federation was still working, performance here and at Kbin lab2 tanked, but interestingly, not Kbin.social.
Looks like I’ll be using Mastodon to interact with the threadiverse for a while. If anyone has any idea what’s going on, please comment.
Thanks for the poke. I’ll take a look
no idea what’s going on here on fedia… hey @jerry
In case anyone asks, why use Mastodon? Why am I not on Lemmy?
Well, yesterday, when the servers were overloaded, I had another user leak through the web interface on lemmy.world. 0.18.1 RC7. It’s not just websockets, even though it fixed a lot. And also, I don’t want to overload Kbin’s servers.
You can go ahead and add “systemic XSS vulnerabilities” to why I’m not on Lemmy, what the heck.
In addition to the sidebar and, potentially, the markdown interpreter, archiving a Lemmy post and accessing it changes the domains to web.archive.org. Yikes.
Lemmy is probably gonna need an audit at this point.
Add can’t log out to the list, still unfixed FFS. So even if you know your cookies can be stolen, good luck trying to stop them from being used.
Credit: lemmy [dot] world/comment/1071591
Edit: Oh, in another issue, someone else last week fixed the part where error pages show your tokens. An audit is definitely in order.
deleted by creator
11:16 PM UTC: Lots of 500 errors on new federated posts. Looks like we’ll be waiting a while.
3:09 AM UTC: Yeah, these 500 errors aren’t going away anytime soon. They stopped for a brief moment while with some fixes, then came right back. @jerry thinks it might be something in the updated build we’re running. I noticed CSS updates that were newer than readit.buzz…
Speaking of which, I wonder why readit.buzz isn’t falling over. It’s about as active as we are right now, and it’s still federating. Maybe 30000 extra posts is the limit? Or maybe there’s more corruption somewhere in Fedia’s servers? IDK
Yes, 500 errors seem to have increased again, struggling to even get thread views (turning off magazine icons, avatars, and sometimes thumbnails helps), but a lot of posts themselves aren’t displaying on their own. This one doesn’t work at all:
https://fedia.io/m/fediverse@lemmy.world/t/66964/If-Meta-federates-with-the-Fediverse-do-my-Mastodon-posts
Some others I can eventually get to display just by toggling thumbnails off and on several times. That definitely feels like a #kbin bug rather than anything to do with #Fedia.[edit] This is an odd one, I managed to get it to display by toggling thumbnails on and off, then after I’ve toggled them again it now appears to be permanently broken: https://fedia.io/m/fediverse@lemmy.world/t/66849/Federation-Lag-o-meter
3:51 PM+1 UTC: Lab2 is down. Looks like ernest@kbin.social is fixing something…
5:40 PM+1 UTC: I replied to the wrong comment about how commenting on how people commenting on this post has caused the post itself to 500, maybe. Deleted; another 500. Did anyone see it? Well in any case, I won’t be able to see your response for a while.
I’m seeing a lot of 500 errors on posts too.
This might be a factor: https://fedia.io/m/fediverse@kbin.social/t/64375/Lemmy-apparently-hit-with-another-bot-wave-tonight-623686-overall
I’m not sure if this is just a me thing or consistent, but i’ve been seeing a lot of 500 Errors lately, even on local magazines.
I’m seeing the same
the image issue seems to be getting worse. I am working on a bandaid to keep it fixed.
Saying hi from kbin.social as a test for ya… good luck!
works