Hours wouldn’t raise my attention. I’m talking days.
I think they might be using some weird shit for “last mile”.
Given that lemmy.world uses Cloudflare, I wonder if their server is overloaded enough to sometimes exceed Cloudflare’s 100 second timeout, triggering http-524 errors during federation.
mox@lemmy.sdf.org 6 months ago
Looks like it received this one, but not the ones I mentioned that are in lemmy.world communities.
jawa21@lemmy.sdf.org 6 months ago
From my experience, lemmy.word can take hours to federate to no matter the instance. I think they might be using some weird shit for “last mile”.
user224@lemmy.sdf.org 5 months ago
From experience sh.itjust.works sometimes has this problem as well.
They also have one thing in common: Cloudflare
Perhaps it’s getting “protected” from federation considered a DoS.
mox@lemmy.sdf.org 6 months ago
Hours wouldn’t raise my attention. I’m talking days.
Given that lemmy.world uses Cloudflare, I wonder if their server is overloaded enough to sometimes exceed Cloudflare’s 100 second timeout, triggering http-524 errors during federation.
mox@lemmy.sdf.org 6 months ago
Would a federation problem on lemmy.world block propagation to slrpnk.net? That would be a weirdly inefficient network design.
That’s an example of what I’m seeing right now. It’s abnormal.
goferking0@lemmy.sdf.org 6 months ago
Can’t wait to see what happens when/if they move to sublinks. How much worse can they get
mox@lemmy.sdf.org 6 months ago
I think kbin.social showed us that overload can get much worse.