Can anybody explain why a long running read-query on secondary replica can block the ghost records cleanup on the primary replica? Thanks.
↧