AG backup preference.

[KeithCorser]

We are currently using the latest version of the backup scripts are running several jobs in parallel to do log backups on hundreds of databases per server. RedGate tools are reporting the following query as high cpu consistently:

SELECT @CurrentAvailabilityGroup = [name],
@CurrentAvailabilityGroupBackupPreference = UPPER(automated_backup_preference_desc)
FROM sys.availability_groups
WHERE group_id = @CurrentAvailabilityGroupID

This is the query behind the function sys.fn_hadr_backup_is_preferred_replica. Can this function call be moved out of the database loop so that it is only called once per job per backup execution reducing the cpu hit?

SQL Server version and edition
SQL 2016sp3

Version of the script
2 Jan 2022

What command are you executing?
Backup log

What output are you getting?
Completes successfully but at high cpu cost.