2 Pitfalls in Priority Probes: Letting One Real Request Through an Open Circuit Breaker
A circuit breaker that stops everything is easy to reason about — until the one job that can't afford to wait gets stopped along with everything else. In my setup, that job was posting: it fell to a 43% execution rate (231 runs vs. 308 skips) because engagement jobs had burned through the quota first. This post is about the fix inside claude-quota-guard.py — claim_priority_probe and run_job ,…
Two common problems arise when using priority probes to allow one requested job through an open circuit breaker. A circuit breaker stops all jobs, but certain jobs, such as posting, cannot afford to wait and must get through. In this scenario, the posting job fell to a 43% execution rate because engagement jobs had consumed the quota first.
To address this issue, the script claude-quota-guard.py was updated with the claim_priority_probe and run_job functions. These functions allow a single priority job to send one real request at a fixed interval, even while the circuit is open. However, there were two pitfalls encountered during this process. The first issue was that the circuit stops all jobs uniformly, but priority jobs should be given a chance to attempt.
The second pitfall was that the last_priority_probe timestamp lives per circuit, not per job, so it doesn't matter which job triggered the probe; it only checks if it has been 30 minutes since anyone last probed.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
