Errors / checked October 9, 2026
Error 1102: Worker exceeded resource limits
The error
A Cloudflare Worker that ran out of CPU time (or memory), often a scheduled Worker started by a Cron Trigger. The dashboard lists it as Exceeded CPU Time Limits.
Quick fix
- Check Workers Logs for the invocation: CPU time and wall time are shown per run. Waiting on fetch(), KV or database calls does not count as CPU time, so the cost is your own code.
- On Workers Paid, raise the CPU limit in Wrangler with limits.cpu_ms. HTTP requests can go up to 5 minutes; a Cron Trigger that runs more often than once an hour is held to 30 seconds.
- Do less per run: process a batch and return, and let the next run continue where this one stopped.
Why this happens
Cloudflare caps CPU time per invocation and returns Error 1102 when a Worker goes over. For Cron Triggers the cap depends on the plan and the schedule: 10 ms on Workers Free; on Workers Paid, 30 seconds for schedules under an hour apart and 15 minutes for hourly or slower.
Other solutions
- Expose the job as an HTTP route on the Worker and call it from an outside scheduler instead of a Cron Trigger. HTTP invocations can use the higher CPU limit and have no duration limit. RunLater calls the route on your cron schedule, retries a failed call and logs each response.
- Move heavy work to a Queue consumer or a Durable Object, or split it across several smaller runs.
Check a schedule before you deploy it: the cron expression explainer shows the next 10 run times.
Frequently asked questions
Does waiting on fetch() count toward the CPU limit?
No. CPU time only counts time spent running your code. Waiting on network requests, KV reads or database queries is wall time, not CPU time.
What is the CPU limit for a Cron Trigger?
On Workers Free, 10 ms. On Workers Paid, 30 seconds when the schedule runs more than once an hour, and 15 minutes when it runs hourly or less often. Wall time for a Cron Trigger is capped at 15 minutes.
Will a bigger CPU limit fix it?
Only if the work fits in it. A job that grows with your data will hit any fixed limit eventually, so batching across runs is the lasting fix.
Have the URL called on schedule, with retries
RunLater calls your URL on a cron schedule or at an exact time, in your timezone, retries failed calls with backoff and logs each response, so a timeout or a 5xx shows up in a log instead of disappearing.