Getting tons of server errors from Google Sheets API
Our application has been hit with massive amounts of 500 errors when trying to access Google Sheets API over the last few days. We’re talking about hundreds of thousands of failed requests.
The error message we keep seeing is:
Sorry, there was a server error. Please wait a moment and try refreshing your document.
What we’ve noticed:
Most failures happen on GET requests to the spreadsheets feed endpoint
This problem started about a week ago
It affects multiple user accounts, not just one specific user
Around 8% of our API calls are returning 5xx errors
The rest return normal 200/304 status codes as expected
Has anyone else been experiencing similar issues with Google Sheets API lately? We can provide detailed logs if any Google engineers are monitoring this forum.
same here! we r getting hit with 500 errors too. thought it was just us! we reached out to google support but no response yet. def seems like a bigger issue on their end.
Same thing happened to us around that time. Our monitoring dashboard went crazy - error rate shot from under 1% to almost 12%. Exact same pattern too - mostly GET requests throwing generic 500 errors while everything else worked fine. When we dug through the logs, we found the errors weren’t hitting all spreadsheets equally. Bigger files with complex formulas got hit way more. We threw together a quick fix by switching to v4 API endpoints where we could and shrinking our request sizes. That got us down to about 3% errors, but it’s still higher than usual. Google’s status page still shows nothing, which is pretty concerning since this seems to be hitting a lot of people.
Same here - started hitting this around the same time. We saw a massive jump in 500 errors from the Sheets API, mostly on batch reads. Fixed it by adding exponential backoff with jitter and cutting our concurrent requests by 40%. The errors are random, not consistent, so it’s definitely on Google’s end - not a quota or auth issue. We also noticed they spike during peak hours, so we moved non-critical stuff to off-peak times and that cut our failure rate way down.
we reached out to google support but no response yet. def seems like a bigger issue on their end.