Retry ranges

Retry ranges

The Run again button setting controls which HTTP status ranges show the Run again button in delivery logs. Multi-toggle — any combination is valid. Tenant-wide.

This is a UI-level filter, not a delivery-level retry policy. Webhooks Pro never auto-retries; Run again is always manual. This setting just decides which rows are eligible for a manual replay.


Where to set it

Webhooks Pro → Settings → Webhooks → Run again button.

Range

Typical meaning

Default

Range

Typical meaning

Default

2xx Success

Successful deliveries

✅ on

3xx Redirect

Redirect responses

✅ on

4xx Client Error

Receiver rejected the request (bad body, auth, etc.)

✅ on

5xx Server Error

Receiver had an internal problem

✅ on

Network Error

No response at all (timeout, DNS, refused)

✅ on

Other / Custom

Anything outside standard ranges

✅ on


What enabling/disabling each range does

When a range is enabled, rows in that range show:

  • The single-row Run again button in the Action column.

  • Are selectable in the multi-select bulk-replay flow.

When a range is disabled, rows in that range:

  • Still appear in the log normally.

  • Just don't show the Run again button.

  • Aren't included when you use the header checkbox to select all eligible rows.

The bulk-replay multi-select on the Logs page respects this setting — header checkbox selects only eligible rows.


Common configurations

Default — everything replayable

All ranges on. Maximum flexibility; admins can replay anything.

Conservative — block 4xx replay

Range

Setting

Range

Setting

2xx

on

3xx

on

4xx

off

5xx

on

Network Error

on

Use when 4xx errors usually mean a config problem (wrong header, malformed payload) that needs a fix on the webhook before any retry. Removing the Run again button on 4xx prevents admins from spamming the receiver with the same broken request.

Network-only

Range

Setting

Range

Setting

2xx

off

3xx

off

4xx

off

5xx

off

Network Error

on

Use when only transient network failures should be retried, and any HTTP-level response — even 5xx — is treated as authoritative.

Read-only audit

Range

Setting

Range

Setting

All

off

No row is replayable. Useful if Webhooks Pro is being used as a write-once audit pipe and any replay would create a duplicate event downstream. Replays would still be possible by toggling a range back on temporarily.


Why no auto-retry?

Webhooks Pro deliberately does not auto-retry failed deliveries. Each Jira event produces exactly one delivery attempt; the result is logged and that's it. Reasoning:

  • Predictability. You know exactly how many times the receiver was hit.

  • No silent doubling. A flaky receiver doesn't get amplified by retries.

  • Manual replay is in your hands. The Logs page gives you a one-click Run again — you decide what's worth retrying after diagnosing why it failed.

If you want auto-retry semantics, build idempotency into the receiver (acknowledge fast, queue for processing) rather than relying on the sender to retry. See Retries and auto-disable for the full philosophy.


Common mistakes

Mistake

Effect

Fix

Mistake

Effect

Fix

Disable 4xx, then wonder why bulk-replay won't pick them up

Header checkbox respects this setting

Re-enable the range for that one investigation, then turn it off again.

Re-enable a range to bulk-replay an outage but forget to turn it back off

Future admins replay 4xx without thinking

Either accept the looser policy or set a calendar reminder.

Confuse this with auto-retry

Expect failed deliveries to retry on their own

They don't. Run again is always manual.


Related