Hello,
Since Friday 13th or Saturday 14th functions doing fetch timeout without clear reason from the user point of view.
This has been partially acknowledged on the status page but the problem is still present and it started long time before
We can see between the working builds and the not working ones because the function size is 0.9Mo heavier.
Can you investigate what is going on with this please ?
Thank you a lot for your help!
Hi, @marcsahuguet! Welcome to the Vercel Community 
Sorry to hear that you’re still seeing issues. Could you share what errors you’re seeing?
Hello @pawlean,
Thank you 
What we will see from the browser perspective is a function timeout
However the problem is not that the code is taking >10 seconds to execute.
To be really sure of this we added a lot of logging and proxied the request going out in the fetch.
However sometimes no logging happens (could reproduce on the first try thought as it is becoming more and more common)
And the proxy shows that sometimes no query is sent sometimes the connection is closed just after the request was made.
Can you look into this please ?
It might be breaking all the deployments doing queries to external APIs which I expect to be a big proportion of all Vercel deployments.
Thank you a lot for your help!
To give more context if you take a request like dwz6v-1726578348640-741d878ac888
It is a really interesting case were the code can start still (which is not always the case) but the handler and console log is called after 9 seconds!
Meaning the current “user function timeout” is ~1 second.
One last data point: try to do a new build of code of the 29th of August were we had no timeout (not a single one) => The issue is there.
How did I do this experiment ? I created a branch on this commit and added a small change to a string with no real impact so that it would need to be a new build.
Thanks for sharing the error code. The incident report you shared was related to Edge Functions, but the error your project is throwing relates to Serverless Functions specifically. You can find more information about possible causes and solutions here: FUNCTION_INVOCATION_TIMEOUT
Hello @amyegan ,
Indeed I do not have data to know whether or not the two issues are linked beside the timing of the two issues
Before posting I went to read the documentation, the code is running fast locally and it is not an edge case described in the documentation.
The engineering team is currently investigating the problem.
It appears that there might be issues using Sentry on Vercel infrastructure currently.
That would not explain all the discrepancies but it is a start.
I will keep you posted on what the engineering team discovers if it can help other users.
Best regards
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.
