We recently updated to 0.18.0 and I haven’t really had a chance to fully test it. I took a snapshot before the upgrade, so we have a working state to roll back to if need be.
Please do not hesitate to post here. Rolling back snapshots will cause us to lose new posts that took place after I took the snapshot.
I saw it running slow on some images, but then realized the image was a 4k image. I have had orange json toast before, and it led to a few seconds of the /error page
Yeah, that’s the sucky part about hosting your stuff internationally. Latency and throughput are severely diminished at certain times because of the vast amount of hops packets have to route through. I’ve got Varnish (cache server) running on the VPS which should mitigate “some” of it as far as images go. But that’s moreso to keep the reverse proxy on the VPS from having to request images all the way from my server box, thereby reducing latency some.
That being said, I’m glad the error page worked. It was @Disabled@burggit.moe 's idea. He coded up the html page (since I don’t know HTML at all) and I got it implemented on the VPS’s reverse proxy to intercept 500/502 errors and redirect to it.