Sudden appearance of "Worker exceeded CPU time limit" errors
Previously-working workers being tested with wrangler dev --env development --remote started failing on Friday, and I can't figure out why. Our production workers appear to still be working fine, but testing workers with wrangler dev is not. The workers will respond to a few requests and then fail with X [ERROR] Error: Worker exceeded CPU time limit. errors. Once they fail with this error, they continue to fail for all requests until restarted. Bizarrely, if I pull from our main branch (which is deployed to production) and run wrangler dev, I still get these errors even though the worker is functioning just fine in production.
11 Replies
How much CPU time is the production Worker using?
How would I test that? Using the unbound workers plan
I see in the workers dashboard our production environment doesn’t have a limit, but the development environment does. There shouldn’t be a limit because we’re on the unbound plan, so I think when accounts were “automatically transitioned” on March 1, something was done incorrectly with that environment
It also seems to be counting invocation runtime incorrectly when using wrangler dev.
Once the worker is started, all requests count against the cpu limits rather than each request having its own cpu limit
The same has happened with one of our production sites running on Pages since this morning.
Check the settings for the worker in your cloudflare dashboard. I found that the migration to the standard usage model set a 50 ms cpu limit on one of our environments that I’m unable to change
The dashboard shows an error saying that I can’t change it because the account isn’t on the unbound workers plan. Which, duh, that plan was phased out on March 1
I recommend opening a support ticket with CF, that’s what I’m doing
Yes, please open a ticket. If you can give the ticket # here as well it will be helpful for escalation/gathering these.
It's not just you that is being impacted/others have been asked to open tickets as well to track the issue
I would watch this incident for further updates on the issue, and try again when it is resolved:
https://www.cloudflarestatus.com/incidents/3q7sg3tpp7qv
Thanks. My ticket number is #3169997
Thanks I forwarded up to make sure it was linked to that incident
is this still the case @Razzmatazz ?
nope, it has been resolved with the fix CF deployed
Nice; thank you for responding 😌