These error 500s are driving me nuts. In the past day I can no longer visit my own profile because of the #error500. I can see my notifications, but I’ve become completely unable to continue a lemmy.world thread for which I have replies & need to reply back to.
#kbinBug #kbin #bug #fedia
They’re very random. One of my magazines I can no longer access @ukdtt but I’m not sure if it was caused by something I did, and microblogs aren’t showing on a lot of occasions. @jerry does the database need another kick?
Quite often I’ve noticed for threads if you toggle the image related settings they start appearing. Sometimes even just refreshing the page clears the error.
@chris
@ukdtt @ciferecaNinjo
I’m giving in and adding a cron job to run a set of cleanup queries every 30 minutes. I’ll have that in place in a few hours. I’m really sorry for the problems.
No worries, appreciate you running this instance, and given how new it is issues are inevitable.
@jerry I appreciate the effort but I must say it’s doing nothing for the error 500 when I attempt to visit my profile. It’s been a continuous error 500 for the past 2 days.
I heard #LemmyWorld has been under a javascript injection attack. The thread I started there (from here) never gets past the spinny mahoochy:
https://lemmy.world/post/1196255
I always have to force reload to view that page. I mention this because my profile page likely tries to access posts from that thread. So I wonder if the error 500 might be caused by timeouts on external services. Just a guess.
@ukdtt @chris
I think these 500 errors are bugs in #kbin. Whilst they might be caused by something in the database, the code ought to be fixed to fail gracefully. It’s possible that some are timeout related, maybe the location the media files are isn’t responding quickly (I’m not sure if they are local files). I know that if I refresh some threads the page displays and it doesn’t appear that anything on the sidebar changes. Maybe that points to a timeout issue too? When refreshed whatever it was fetching is now in the cache so it works.
It probably needs somebody familiar with the code to look at the error logs to figure out what is going on.