AG unhealthy node auto evict
Idea is to find the disconnected replicas and evict them before the primary replica tlog fill-up.
When a SQL Server secondary replica in an AlwaysOn Availability Group becomes unavailable or unhealthy, the SQL Server transactional log residing on the primary replica is not committed. The primary replica continues to accumulate transaction logs until the unhealthy or unavailable replica becomes available in a healthy state to resume. The primary replica AG transactional log file will fill the available disk storage if the unhealthy replica issue is not resolved within a reasonable time or if the unhealthy replica is not evicted from the availability group. Unavailable disk space to process new transactions will result in database outage to applications.
The customer approached Premier Services Designated Support Engineer (DSE-PFE) to address the issue and to resolve the problem.
This solution is already implemented in Cx systems. It is good to add it as a feature in the SQL product.
Contact me for further details.