add config to adjust the changefeed error stuck duration #9875
Closed
Description
Is your feature request related to a problem?
In certain situations, the CheckpointTs of a changefeed may get stuck due to a slow table for a period exceeding 30 minutes. In such cases, if the sink module reports any warning, the changefeed will immediately transition to a failed state. At this point, users must manually restore the changefeed, and data from other tables also needs to be resynchronized starting from the CheckpointTs. This can be quite inconvenient and unnecessary.
Describe the feature you'd like
Add a configuration option that allows controlling the duration for which a changefeed does not transition to a "failed" state when it gets stuck due to an error or due to a slow table.
Describe alternatives you've considered
No response
Teachability, Documentation, Adoption, Migration Strategy
No response