Open thewh1teagle opened 1 year ago
After many hours of trying to find why my app isn't working, I found that peerjs default server was down.
My solution was to run peerjs server for free using fly.io (constant free for side projects)
I made simple docker file of peerjs server, configured fly.toml limits to something high and it works fine.
Hm. That would be terrible if true, but I'm not convinced yet this is the case. Starting 12 hours ago (that might be yesterday, depending on your time zone), the status page shows requests intermittently timing out. You might have experienced them.
The monitoring is from 9 different locations checking for status code and response content; I’m quite confident it works reliably. Nonetheless I started monitoring the server with another monitoring service. We’ll know more if the reported uptimes differ. I’ll also try to increase the sensitivity of the timeouts. Maybe it timed out in the browser while the monitoring waited long enough.
We also gather logs on the CDN side. Having origin timeouts at all isn’t something that happens regularly, so this is not happening often. But thank you bringing this to our attention.
For debugging: ping
will only reach the CDN, so it will always be up. wscat
will always return 403, as it is filtered in the WAF.
https://github.com/peers/peerjs/issues/809#issuecomment-1818027729
Yesterday the server was down and the status page showed it was up
So it's possible that it happens many times and status page doesn't catch it / log it