Incident Dossier

Retry Budget Shared Across Callers

After deploying a new rate-limited upstream integration, the team wired up a shared retry budget to protect the upstream. Within an hour, calls from an unrelated service started hitting BudgetExceeded even though the upstream was healthy and individual callers were succeeding. The budget counter was being driven to zero by callers that had long since completed.

Initial failing signalSuccessful callers leave the budget depletedExpected behavior: After a successful caller uses retries, the shared budget returns to its starting value. · actual behavior: The shared budget remaining is 0 after a successful caller uses retries.

Progress

1Reproduce the bugRun checks

2Edit the fixOpen

3Pass checksOpen

4Submit fixOpen

Reproduction Steps

Run npm test.
Inspect withBudgetRetry and follow what happens to the consumed slot after the call resolves.
Notice that the helper never calls budget.refund, even on the success path.
Track how many slots each call consumed and refund them before returning or re-throwing.

Focus Mode

Start

src/budget.js

Likely area

src/budget.js

Challenge Coach

Hints reveal broad direction first, then narrow toward the suspect path without showing hidden tests or exact solution code.

TopicReliability

DifficultyMedium

Est. time22 min

Workspacepreparing

Explorer

Loading files

Evidence Map

Failing signal loaded from challenge API
Workspace files loaded from workspace API
Results rendered only after API run/submit

Autosaved

workspace

Outputread-onlyready$ prepare workspace

Preparing workspace