Gloss Tools

Retry Time Budget Planner

See which serial retry attempts fit a time budget when each attempt uses its full timeout and delays have a cap. This is a time model; it never issues retries.

Total budget and per-attempt timeout

Integer, 0 to 86,400,000 milliseconds.

Integer, 1 to 600,000 milliseconds. Each attempt is assumed to use its full timeout.

Attempt limit, initial delay, multiplier and delay cap

Integer, 1 to 16 attempts.

Integer, 0 to 600,000 milliseconds. There is no delay before the first attempt; delays start from the second attempt.

Integer, 1 to 8. Each delay is multiplied by this factor, up to the delay cap.

Integer, 0 to 600,000 milliseconds. When the initial delay exceeds the cap, the cap applies from the second attempt.

Nothing calculated yet. Enter values or load the example.

How it is calculated, and the report format

The first attempt has no delay: it starts at 0 ms and ends at the per-attempt timeout. For attempt i (i ≥ 2), the delay is the initial delay multiplied by the factor to the power of i−2, capped at the delay cap; the delay is added after the previous attempt ends. The first attempt whose end exceeds the total budget is not kept, and its waiting time is not deducted; an attempt ending exactly at the budget fits. Unspent time is the budget minus the end of the last kept attempt, or 0 when nothing was kept.

Every field accepts ASCII decimal integer strings only, with no spaces, signs, decimals, or exponents, and at most 8 digits per field. The downloaded report is always named retry-budget-report.json and contains the result as compact JSON, with keys sorted by UTF-16 code unit order and one trailing newline.

Frequently asked questions

Does an exact fit count?
An attempt ending exactly at the budget fits.
Does it issue retries?
It calculates a timeline assuming every attempt uses its full timeout.

All calculation happens locally on this page. Nothing is uploaded, stored, or logged.