BE 2010 R3 - B2D Concurrent and Consecutive Jobs

[craig_mcc]

Hi all,

I would like some help or advice about a strange situation that is occurring with a daily rotation of multiple B2D jobs configured as separate jobs, with separate selection lists via a policy.

I have followed the Symantec best practices advice and configured a GFS rotation with a SATA HDD for each day (Mon-Thurs), 3 weekly drives (Weekly1-3) and 3 monthly drives (Monthly1-3).

Each separate drive has it's own B2D folder and each of these folders have been added to the corresponding pool (i.e. Daily Pool, Weekly Pool, Monthly Pool), which the jobs reference.

I have set the number of concurrent jobs to 1 on both the device and the general preferences, however, no matter what I do, 2 jobs run concurrently every night, which breaches our backup window due to some overlap in processes running.

If the jobs are run consecutively, we make the window and the throughputs are double if we run concurrently.

Can anyone please shed some light on why 2 jobs are running concurrently, even though I have selected that only 1 run?

Thanking in advance,

Craig.

[ArulPrasad]

Hi ,

The Scenario explained Needs More details About Number of b2d devices connected to Each Pool.

And Number of Selection List assigned to the policy (i.e Number of Daily Jobs etc..)

Rember That If the Job is targeted Direct to the device with concurrent operation set to 1 and it shuld work as expected.

If its Targeted to a Device pool (Contating more than one device)

its obivious second Job will run.

Concurrent operation is for Device(b2d) not for the job or Device pool.

Regards

Arul

[pkh]

The concurrent job is not per pool, but per B2D folder.  When you target the jobs to the pool, each job will pick a device from the pool.  If you have 2 devices in the pool that is on-line, then 2 jobs will run run simultaneously.

[craig_mcc]

Thank you for the prompt replies.

#apr

I have 4 daily B2D folders in the Daily Pool, 3 weekly B2D folders in the Weekly Pool and 3 monthly B2D folders in the Monthly Pool.

Each of the physical HDD's contain the appropriately named physical folder - i.e. E:\Wednesday, E:\Weekly1, etc. for each of the HDD's and folders. Each HDD only has 1 folder, relating to the day of the backup.

Only 1 physical HDD (and therefore B2D folder) is ever connected to the system and online.

#pkh

I understand what you are saying about the B2D folder being in control of the number of concurrent jobs, however, even if there are 4 B2D folders in a pool, the set of jobs only ever utilise 1 folder (the day of the week), so I woulld assume that it should only ever run 1 concurrent job as set by the options on the B2D folder.

Am I missing something?

Thanks,

Craig.

[pkh]

Let's say you have B2D-1, B2D-2, B2D-3 and B2D-4 folders in your device pool and are all on-line.  When you run 4 jobs targeting the pool, then each job will use 1 of the folders.  For exampl,e Job 1 may use B2D-1 , Job 2 may use B2D-2 and so on.

If the jobs above target B2D-1, then they would queue up one behind the other.

When you target a device pool, any on-line device in the pool is fair game.

[Colin_Weaver]

Are you aware of the known issue with B2D folders being created on the wrong disks if you restart the BE services

i.e. you created a Daily Device Pool using 4 physical USB disks with one unique B2D device named on each one, but these disks usually connect to the operating system on the same drive letter, then a service restart will cause the B2D devices form the other USB disks to be accidentally ceated on whichever disk you have inserted at the time.

http://www.symantec.com/docs/TECH155949

Basically if all is working correctly then if your Monday disk is plugged in but the other disks are not then ONLY the Monday B2D device should show online in devices. If the other B2D devices show as online that will be part of the problem