Retry Backoff Schedule Simulator
Expand a retry policy (exponential / linear / fixed, factor, cap, jitter, attempt budget) into a per-attempt delay and cumulative wait schedule — jitter as a min/expected/max range instead of one sampled draw, plus a verdict on whether the cumulative wait overruns your own timeout
forge-retry-backoff-schedulev1.0.0AWS Architecture Blog "Exponential Backoff And Jitter" full/equal jitter pair (full = random(0,d); equal = d/2 + random(0,d/2)), as cited by k-lab.dev/retry's FAQ per brief §7 know-how #3 · retries.dev "Max Retries (excludes initial request)" attempt convention per know-how #1
Workspace
Local or server · same path as API
dev/retry-backoff-scheduleMultiplier applied per retry. Must be greater than 1.
Leave empty for no cap. Once the cap is reached, every later delay stays flat at it.
1 to 100.
Optional. Leave empty to skip the budget check.
On: the number above counts retries only, the initial request is extra. Off: it counts every attempt, initial request included.
Fix the highlighted field to see the schedule.
Computed from the policy alone — nothing is sent to any upstream service. Jitter is reported as the range a real client falls inside, never as one sampled draw.
API
forge.dev.retry_backoff_scheduleSame contract as the page. Use an authenticated key in production.
curl -X POST 'https://forge.nebutra.com/v1/tools/dev/retry-backoff-schedule/invoke' \
-H 'Authorization: Bearer sk-sailor-…' \
-H 'Content-Type: application/json' \
-d '{
"example": true
}'