Backup Exec 11d installation - need some basic guidance

[bchesley]

I inherited an installation at a client and even after scanning the admin manual's 1600+ pages can't seem to get things working based on what a prior party setup.  Basic config is the software is installed on a Windows SBS2003 server and using M-F external USB drives (600G capacity).  There is only one job set up and set to append then overwrite with no overwrite or append horizon set (i.e., 0 hours retention).  See sections of job summary below:

The device is - Removeable Backup To Disk FolderWhen this job begins - overwrite media
Type - Full backup using archive bit

My problem is that the job is constantly failing when running out of disk space on the external drive (i.e., shouldn't the set be pruning older copies by overwriting with the newer set?).  The full backup results in around 200G of backup content.  I can clear space in explorer and get the job to run but then the catalog is no longer accurate.

Thanks in advance for any suggestions.  This shouldn't be this hard me thinks.

bc

PS - below is some output from emailed notification log on failed job for one image:

Backup set #2 on storage media #20

Drive and media mount requested: 3/22/2011 10:40:26 AM

Drive and media information from media mount: 3/22/2011 10:40:27 AM
Drive Name: Removable Backup-to-Disk Folder
Media Label: B2D000918
Media GUID: {95aaccd5-377a-454b-92d1-
				
b7768c95802b}
Overwrite Protected Until: 12/31/1899 7:00:00 PM
Appendable Until: 12/31/1899 7:00:00 PM
Targeted Media Set Name: Backup
				

[Ken_Putnam · Accepted solution]

There is only one job set up and set to append then overwrite with no overwrite or append horizon set

Looks like you are doing most things correct.  Jobs set to overwrite, not append, but I have seen some really strange behavior with Overwrite of 0.  Try setting it to 1 hr and see what happens.  Also, from Tools\Options\Media Management, be sure that you have selected either PARTIAL or FULL .

Selecting the option to "use overwriteable media in the target media set before scratch media" should start overwriting the BKF files when they are past the OPP.  if you choose the other "Use scratch media before...", then Backup Exec will fill the available space on the volume before reusing existing files.   If this is a dedicated backup volume, either option should work OK

One thing I notice is that the device is a Removeable B2D.  Should be a standard (R-B2D is designed for devices where the hardware stays connected and the media is removable, like Jazz or RDX drives.)

Do each of your B2D disks have it's own drive letter, and always mount to that same letter?  Do you inventory the device after (re)connecting it?

[bchesley]

Each external disk mounts to the same drive letter and on the drives there are two subdirectories f:\backup and f:\backup1.  I am attaching a screen shot of the device tab to make this clear.  Also, attaching the options->media management settings FYI.  I have not been inventorying after reconnect as a matter of procedure although I have run that command and not been able to tell what the benefit is (i.e., my view of online media does not seem to change).

I will make the OPP change you suggested and also route the job to the non-removeable device as that makes total sense.

Thanks for the insights.

[Ken_Putnam]

Hmm

prior to 2010R2,  Veritas and then Symantec recommended against shareing a single drive letter between multiple external drives.  (2010R2 has greatly enhanced external drive handling)  The recommended way was to assign each physical drive it's own drive letter, and alway mount each drive to that unique letter. Then place the B2D devices in a pool, point your job(s) to the pool, and let BackupExec determine which foldler to use. 

You have to nudge BE to tell it which drive is currently attached, and can either manually pause each drive before disconnecting it then unpause the drive/folder that you have just connected, or run a CMD file to execute BEMCMD commands to pause all your B2D drives and then unpause them all.  The only one that BE should think is there afterwards should be the one that is physically connected

[bchesley]

Looks like I will get a successful full backup tonight.  Thanks again.