Supabase timing out
Hello, roughly 16 minutes ago our Next.js app's all supabase calls started timing out with no code changes or redeployments.
Is this an issue with Supabase?
32 Replies
Hard to say. Any specific errors? Anything in the API Gateway logs?
Any status issues in your main project page?
lots of 504s, no issues in status page
Are requests getting to the API Gateway logs?
What region? Although not seeing anything today other 2 reports with status like this that resolved quickly.
yeah they are, logs look like this though
surely if it was an issue on my side the errors wouldnt be 504?
So just auth?
looks like it
Try restarting the instance.
Any change to your configuration like going to new API keys or asymmetric JWTs.
See if auth log has any info.
Do you have a support request in?
Any change to your configuration like going to new API keys or asymmetric JWTs.no
See if auth log has any info.context deadline exceeded
Do you have a support request in?no
do you guys even have live chat? i dont see it
Do you have any trigger functions on auth tables or using an auth hook?
Not sure what that is. This is a user helping user forum and not monitored by Supabase.
yep but theyve never been an issue before
yea bet
What do they do?
Do they make external calls? Like to an API or logging platform?
Seems like something in the auth process is timing out.
So not the Gateway to your project, but something about the actual auth requests.
no they just call some sql functions
support replies in 1 business day too apparently lol
nice support considering the platform breaks every other month
Are you pro?
yep
Did you try a restart?
Any postgres log errors from your auth SQL logic?
restart didn't work
out of desperation we upgraded the instance
seeing if that does anything
Any postgres log errors from your auth SQL logic?no and it just creates a row in another table shouldnt cause anything
our platform is down for 2000+ users
so we're willing to go for nuclear option
Your other Database calls seem to be working so not a database timeout in general.
we just realized
our CPU is pinned to 100%
all of a sudden
A restart would normally clear that if it was something locked up.
ya exactly
CPU was pinned before and after
When you say you upgraded what did you upgrade from and to.
nvm issue was on our side
bug in other rogue process which caused insane amount of requests
thanks for the help 🙏
Glad you found it!
L -> XL
Thank you for the help 🙌