Sql-server – T-Log Backup in AOAG

My question is: - Do i have to schedule the jobs at primary replica and then disable them until fail-over occurs and then manually enable jobs or what is the best way to deal with these please?

No, you should just add a check with the new function sys.fn_hadr_backup_is_preferred_replica

if sys.fn_hadr_backup_is_preferred_replica('your db name') <> 1
begin
    print 'This is not preferred backup replica .....'
        -- If this is not the preferred replica, exit (without error).
end
else
begin
    print 'This is the preferred replica .. continue with backup'
        -- If this is the preferred replica, continue to do the backup.
        -- here goes the backup command .....
end

Alternatively, you can use SQL Server Maintenance Solution - by Ola hallengren which supports AlwaysON as well.

As long as you don't care that:

• in the async model, the REDO queue may fall way behind, so the backups on your secondary may not represent the point in time that you think they would if they were taken on the primary.
• you will not be able to use DIFFERENTIAL backups in this model.
• in the event of a restore, you will have to manually determine which log backups contain the LSN ranges you'll need to restore on top of the chosen COPY_ONLY backup and, given the above, the timestamps on the bak/trn files may not be accurate enough to determine the set of files and sequence you need.
• backup history will not be recorded centrally.

Also, I still think there are going to be scenarios where this ultra-long log chain, or relying on a COPY_ONLY backup taken in between log backups, are going to make it difficult to restore. Personally, I would be keen to perform a full, proper backup on the primary on some regular interval, say once a week. YMMV.