How do we monitor the replication status between two cross data center clusters.?
Do we have any table/veiw which actually stores the current replication status of the database I know we have tables where we have data to check ‘is_dr, primary,sec. etc’ but would like to know the status in case of replication has some issues.??
We use the following query to look at our replication status and alert of any backlog:-
select DATABASE_NAME,PRIMARY_URI,PRIMARY_STATE AS p_state,SECONDARY_URI,SECONDARY_STATE AS s_state,LSN_LAG,VOLUME_LAG_MB,REPLICATION_THROUGHPUT_MBPS,ESTIMATED_CATCHUP_TIME_S
from information_schema.mv_aggregated_replication_status
WHERE SECONDARY_URI IS NOT NULL
Internally, we capture these stats to a timeseries to chart out replication speed/lag
Challenging part is will be once the primary is down the mv_aggregated_replication_status table values are gone and becomes difficult to automate the failover from primary to secondary cluster.