Netbackup 6.5 not using one of the drives

[drj003]

We use Netbackup 6.5.  The drives are housed in a tape library.  One out of the 5 drives is not being used.  The drive is up, but will not accept any tapes.  How can I make sure that the drive path is setup correctly?  Thanks.

[drj003]

I just noticed that it's not always the same drive that's not writing.  So basically it's only using 4 out of 5 drives at a time.

[Marianne]

What is the number of drives specified in the Storage Unit properties? (Maximum concurrent write drives  )

Are enough jobs initiated for 5 drives?

Is there enough media available?

[drj003]

There are 2 storage units.  One has 3 Maximum concurrent drives and the other has two max concurrent drives.  Both are apart of a storage unit group.  We have one master/media server with 4 scsi drives attached to it.  We have one media server with 1 scsi drive attached to it.  Is there an issue with that?-  Should there be one storage unit with 4 max concurrent drives, and one with 1, since that's how many drives are attached to each server?

There are more than enough jobs for 5 drives.

There are plenty of unassigned tapes in the volume pool.

Thanks for your reply.

[Marianne]

What is the Media Server/Host Connection set to for these two storage units?

If Media Server for the 3-drive STU is master, then the master will only ever use 3 drives.
If Media Server for the 2-drive STU is media-server, then the media server can only ever use 1 drive.

The Storage Units should reflect Host Connection. If the drives are of the same density and in the same robot (0), you need something like this:

Label: master-hcart2-robot-tld-0
Storage Unit Type: Media Manager
Host Connection: master
Number of Drives: 4
......
Robot Type/Number: TLD / 0
.....

Label: media-server-hcart2-robot-tld-0
Storage Unit Type: Media Manager
Host Connection: media-server
Number of Drives: 1
......
Robot Type/Number: TLD / 0
.....

[drj003]

The storage unit with 3 concurrent drives; the media server was set to "master server".  I increased the number of concurrent drives to 4.

The storage unit with 2 concurrent drives; the media server was set to "any available".  I changed the media server to "media server" and decreased the concurrent drives to 1.

There are still only 4 drives writing.  The one that isn't writing says "TLD" instead of "active" in activity monitor. 

Thanks for your help on this.

[Marianne]

Please post output of 'bpstulist -L'.

I have often seen that changes to Storage Units are not applied right away. I normally issue "bprdreq -rereadconfig" after changing attributes to force a 'config reread'.

[drj003]

Here is the output of bpstulist -L

Thanks

D:\Veritas\NetBackup\bin\admincmd>bpstulist -L

Label:                nad2backupsb-hcart2-robot-tld-0
Storage Unit Type:    Media Manager
Host Connection:      nad2backupsb
Number of Drives:     4
On Demand Only:       no
Density:              hcart2 (14)
Robot Type/Number:    TLD (8) / 0
Max Fragment Size:    1048576
Max MPX/drive:        6

Label:                nad2vmbak-hcart2-robot-tld-0
Storage Unit Type:    Media Manager
Host Connection:      nad2vmbak
Number of Drives:     1
On Demand Only:       no
Density:              hcart2 (14)
Robot Type/Number:    TLD (8) / 0
Max Fragment Size:    1048576
Max MPX/drive:        6

Label:                Puredisk_storage
Storage Unit Type:    Media Manager
Host Connection:      nad2backupsb
Number of Drives:     4
On Demand Only:       no
Density:              hcart2 (14)
Robot Type/Number:    TLD (8) / 0
Max Fragment Size:    1048576
Max MPX/drive:        6

Label:                Hotbackup
Storage Unit Type:    Disk
Media Subtype:        Basic (1)
Host Connection:      nad2vmbak
Concurrent Jobs:      1
On Demand Only:       yes
Path:                 "D:\Catalog\HotCatalog"
Robot Type:           (not robotic)
Max Fragment Size:    524288
Max MPX:              1
Stage data:           no
Block Sharing:        no
File System Export:   no
High Water Mark:      90
Low Water Mark:       70
Ok On Root:           no

Label:                Backups
Storage Unit Type:    Disk
Media Subtype:        Basic (1)
Host Connection:      nad2vmbak
Concurrent Jobs:      1
On Demand Only:       yes
Path:                 "F:\Backups"
Robot Type:           (not robotic)
Max Fragment Size:    524288
Max MPX:              1
Stage data:           no
Block Sharing:        no
File System Export:   no
High Water Mark:      98
Low Water Mark:       80
Ok On Root:           no

D:\Veritas\NetBackup\bin\admincmd>
 

[Marianne]

If NBU is STILL only using 4 drives, please see in Activity Monitor if the Media Server is used at all.

I assume that you have run test-backups to the media server STU to confirm functionality?

How is the STU-group configured? What is the selection order/criteria?

What is the reason listed in Activity Monitor for queued jobs?

[drj003]

It appears that NBU is using the media server tape drive.

The storage unit group has the following 2 storage units-

nad2backupsb-hcart-robot-tld-0

nad2vmbak-hcart2-robot-tld-0

The storage unit selection is set to "Round Robin"

Here is the detailed status of a queued job-

8/1/2011 7:55:34 PM - requesting resource AllTapes
8/1/2011 7:55:34 PM - requesting resource nad2backupsb.NBU_CLIENT.MAXJOBS.scooby
8/1/2011 7:55:34 PM - requesting resource nad2backupsb.NBU_POLICY.MAXJOBS.NASDEV-UNIX
8/1/2011 7:55:34 PM - awaiting resource AllTapes - No drives are available

Here is a detailed status of a job that I believe is using the media server storage unit-

8/2/2011 8:13:13 AM - requesting resource AllTapes
8/2/2011 8:13:13 AM - requesting resource nad2backupsb.NBU_CLIENT.MAXJOBS.PS-SVN01
8/2/2011 8:13:13 AM - requesting resource nad2backupsb.NBU_POLICY.MAXJOBS.SVN-LUNT-Svrs
8/2/2011 8:13:13 AM - awaiting resource AllTapes - No drives are available
8/2/2011 8:13:48 AM - granted resource nad2backupsb.NBU_CLIENT.MAXJOBS.PS-SVN01
8/2/2011 8:13:48 AM - granted resource nad2backupsb.NBU_POLICY.MAXJOBS.SVN-LUNT-Svrs
8/2/2011 8:13:48 AM - granted resource 001355
8/2/2011 8:13:48 AM - granted resource HP.ULTRIUM2-SCSI.000
8/2/2011 8:13:48 AM - granted resource nad2vmbak-hcart2-robot-tld-0
8/2/2011 8:13:49 AM - estimated 130180032 kbytes needed
8/2/2011 8:13:51 AM - started process bpbrm (4584)
8/2/2011 8:14:03 AM - connecting
8/2/2011 8:14:04 AM - connected; connect time: 00:00:01
8/2/2011 8:14:09 AM - mounting 001355

Thanks for you help.  Please let me know if you would like any more information.

[Marianne · Accepted solution]

I am out of ideas... I was convinced that all would be fine once the STU properties were changed.

If there was one particular drive that was not used it could be an orphaned drive allocation but this does not seem to be the case here. Maybe still worth a check?

Please check with 'vmoprcmd -d -h <hostname>' for both server to see which drives are in use.
Then check resource broker drive allocation with 'nbrbutil -dump' and check the MDS allocations at the bottom of the output.  Compare the allocations with what is actually in use from vmoprcmd.

If there should be a descrepancy, you can use -releaseMDS mdsAlocationKey to release.