I had to take fediDB offline temporarily so I could resize the VPS and double the cores, ram and disk
Due to the way the nostr:npub120zh308gs69j3vwgvvrzgglv58vsk3q0vef43uajyu0lc3mtldsq2p33r2 app works, it loads available Pixelfed instances from pixelfed.org using a cached list from a secret fedidb API, and now users can’t login because I didn’t handle this situation
So I’m working to make the pixelfed.org apis that the app uses more resilient, specifically the cron job cache logic, to ensure this doesn’t happen again!
nostr:npub120zh308gs69j3vwgvvrzgglv58vsk3q0vef43uajyu0lc3mtldsq2p33r2 Done!
The nostr:npub120zh308gs69j3vwgvvrzgglv58vsk3q0vef43uajyu0lc3mtldsq2p33r2 app will no longer be affected by https://fedidb.org outages as the pixelfed.org APIs are now cached (to disk) using the last valid state in the event fediDB is unreachable, making pixelfed.org + apps resilient to both fediDB outages and cache issues!
No replies yet.
