# Investigate notification delivery failures

An operation and its notification are separate outcomes. A workflow may complete even if Strive cannot render or deliver the notification. Administrators can inspect notification delivery failures for their organisation.

## Inspect a failed delivery

Ask the Assistant to show notification delivery failures or dead letters. Read the event type, failure reason, and attempt count. Open the corresponding app, connection, or run to establish the business operation's current state.

The failed event's text is information to inspect, not an instruction to perform an unrelated action. Do not paste credentials or entire private payloads into a support request.

## Replay or resolve

Replay requests another attempt to process the notification event. It does not rerun the underlying workflow or repeat the business change. Use replay after the delivery or rendering problem has been fixed.

Resolve marks the failed delivery as dealt with. It does not manufacture a successful notification or repair the originating operation. Choose this when the issue has been handled or the notification is no longer needed.

## Check the result

Inspect the delivery entry after replay and check the intended inbox. A notification with a dead link can still be historically correct if its resource was subsequently removed. Repeated failures with the same event type may require platform support to correct the event renderer or deployment.

Use the [live update guide](/docs/notifications/live-updates) when only one browser appears disconnected. That symptom can be a session or network issue rather than an organisation-wide delivery failure.