Connection recovery & capacity
Recover without duplicating work.
See connection recovery attempts and temporary provider capacity problems.
Recover without duplicating work.
When a request is interrupted or capacity is temporarily unavailable.
Connection errors and capacity limits have different causes. Recovery keeps the interruption visible and uses bounded retry behaviour before leaving work paused.
Follow retry status and countdowns. When uncertainty remains about an external action, inspect the task before resuming.
How to get started.
From your first choice to the result. Here is how to use this feature in your own project.
- 01
Review the reported error, attempt number and any waiting period.
- 02
Allow the existing recovery policy to retry when appropriate for that error.
- 03
If retrying is unsuccessful or unsafe, the task remains paused for later continuation.
How it could work for you.
A short network outage clears and the conversation continues from its recorded state.
How to assess the result
A reconnect does not prove an interrupted external operation had no effect.
Where to find it.
The active task’s status
Available actions depend on your plan, mode and project state. Features that require a task become available within an appropriate conversation.What is saved?
Errors and recovery events remain with the task.
What you need to know
Recovery behaviour depends on the error and engine. Reconnecting must not blindly repeat an external action that already happened.