I'm trying to use TSQL to do AlwaysON health monitoring. After going through the documentation, i found that the health related information
lies in a AlwaysON related DMV's(Dynamic management views). One of the DMV's i'm looking at is "sys.dm_hadr_availability_replica_states". This DMV contains the following 4 useful columns namely:
1) Connected State
2) Operational State
3) Recovery Health
4) Syncronization Health
From what i have absorbed after reading the documentation, "Connected state" on a primary node indicates whether the secondary nodes are connected to the primary node. On a secondary node, it indicates whether secondary is connected to primary.
I'm confused between the other 3 columns i.e. "Operational State", "Recovery Health" and "Syncronization Health". Could i please get some help on how should i interpret these columns on secondary and primary replicas ?
Also, while looking at the "sys.dm_hadr_availability_replica_states" data on a secondary replica, i found that for one of the Availability groups, recovery_health was NULL even though is_loca was "1". Per MS documentation, recovery_health = NULL means is_local = 0. This does not seem to be the case here. synchronization_health for this Availability group was "NOT_HEALTHY". Not sure if this is a bug(documentation/functionality) ?
1) Connected State
2) Operational State
3) Recovery Health
4) Syncronization Health
From what i have absorbed after reading the documentation, "Connected state" on a primary node indicates whether the secondary nodes are connected to the primary node. On a secondary node, it indicates whether secondary is connected to primary.
I'm confused between the other 3 columns i.e. "Operational State", "Recovery Health" and "Syncronization Health". Could i please get some help on how should i interpret these columns on secondary and primary replicas ?
Also, while looking at the "sys.dm_hadr_availability_replica_states" data on a secondary replica, i found that for one of the Availability groups, recovery_health was NULL even though is_loca was "1". Per MS documentation, recovery_health = NULL means is_local = 0. This does not seem to be the case here. synchronization_health for this Availability group was "NOT_HEALTHY". Not sure if this is a bug(documentation/functionality) ?