25112
asked on
do you keep backup job on both primaqry & sec nodes, on HA setup. (SQL Server 2012)
in HA (not synch), you have primary and secondary nodes/instances.
1)for Daily backup jobs (of databases), should you keep identical jobs on both servers? (if one fails scenario). is this best practice?
2)will below idea help to keep the job only one server/node:
If sys.fn_hadr_backup_is_pref erred_repl ica(@dbnam e) <> 1 BEGIN <do not backup> ELSE <do backup> END
1)for Daily backup jobs (of databases), should you keep identical jobs on both servers? (if one fails scenario). is this best practice?
2)will below idea help to keep the job only one server/node:
If sys.fn_hadr_backup_is_pref
ASKER
jmcg, both nodes have same amount of disk capacity (identical)
it seems doing it on secondary is best practice, based on the logical points you mentioned? then, how will you position the job to run and to backup only from secondary? (given that different DBAs might have reason to failover as needed through the week)... so the primary may switch every other day or as long as failovers happen. what is best practice how to keep and maintain jobs on both servers to accomplish this.. to backup only the secondary, judging intelligently which is the secondary at the moment.
it seems doing it on secondary is best practice, based on the logical points you mentioned? then, how will you position the job to run and to backup only from secondary? (given that different DBAs might have reason to failover as needed through the week)... so the primary may switch every other day or as long as failovers happen. what is best practice how to keep and maintain jobs on both servers to accomplish this.. to backup only the secondary, judging intelligently which is the secondary at the moment.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
that helped, thanks.
Presumably, it would be better to run backup on the secondary, since the primary is the one servicing database access traffic, but if one instance has generally better access to the backup storage, that instance should be the one to back up.