Scheduled backups not starting ontime

[somesh_p]

I've a requirement in netbackup that I have a policy with two clients. I want the backup to start for both clients to start at the sametime. Those two are HA clients. Either the backup should start at the sametime for both nodes or shouldn't start at all. Sometime, manual firing of that policy (or) failed backup for one of the node is deviating the scheduled start of that policy. Policy has calendar based scheduled.

Manual backup submission or any other activity shouldn't impact the scheduled start time of policy (for both clients). Thats my requirement. Please suggest how to achieve it

[Dyneshia]

Do you have " Allowed to retry after run day" selected ?

[Will_Restore]

If you are backing up both at same time and HA event occurs, how do you handle that?  I'm thinking it's better to back them up at different times.

[somesh_p]

I dont have "allowed retry after run day" selected.  We have designed a nebackup solution that works for HA cluster. So, failover is handled aptly. My only concern is that HA backup (both clients) is very time sensitive. Backup for both client should start at the same time. sometimes, manaul run or some failures of HA backup, is deviatting the scheduled start time of backup for two clients. That is causing a problem for our HA solution to work correct. Please advise. Regardless what, backup for both clients should start at the same scheduled start time

[Dyneshia]

What version of Netbackup are you running ?

[somesh_p]

version 7.0.1

[Dyneshia]

And OS ?

[somesh_p]

OS is Solaris 10

[Mark_Solutions]

It sounds like manually running backups or failures causing schedule creep is you issue - though not normally a problem with calendar based scheduling.

It may be that you need a very tight backup and frequency window.

By that i mean that you use frequency based scheduling with your start window set to say 4 hours and your frequency set to every 5 hours.

So if you want them to start at 10pm you set a start window of 10pm to 2am (allows for re-runs if they do fail) and a frequency of 5 hours.

This means you can manually run the backup right up to 5pm that day without affecting it re-running promptly at 10pm.

Hope this makes sense and helps

[somesh_p]

I already set up schedule calander based. still, the scheduled start time is deviating from its scheduled start after manual runs and failed backups. We have backup preparation script (bpstart_notify) that shall run simultaneously on both nodes of HA at the sametime. Thats why the scheduled start time is very time sensitive for me. Please suggest me some solution.

Further details, policy is setup with two clients (HA nodes), calendar based scheduling, scheduled start is 00:01 AM and backup window is 15 hrs.

[Will_Restore]

In this case, you are better to use an external scheduler instead relying on NetBackup calendar. 

[somesh_p]

Good idea though to use external scheduler. But, I dont have flexibility to use externa scheduler

[Mark_Solutions · Accepted solution]

I suggested going from calendar to frequency based scheduling .....

00:01 is an odd start time - usually goes in 10 minute increments and NetBackup itself only fires off jobs on the 10 minute mark so they would most likely actually start at 00:10

15 hours is a very long start window - it is only a start window and not a run window

Maybe try my idea for a while - frequency based, starts at 00:10 ends at 04:00, frequency of every 5 hours and see if that does the job for you