look for the bpcoverage command that is in the admincmd dir

you can run this for your server  and it will tell you if a fs is covered and by what policy.

> Single threaded backup set to "All_Local_Drives".  Cross Mount Points and Follow NFS are checked.

Why do you enable Cross Mount Points and Follow NFS with ALL_LOCAL_DRIVES directive?
No need to enable both because ALL_LOCAL_DRIVES covers all local-mounted file systems by itself.
Clear both and try again.

but mounted filesystem is covered in other part of backup. ALL_LOCAL_DRIVES indicate that this backup will take all local filesystem.
/home/db2 is skipped while the job taking data in /home, but later, /home/db2 will be backed up.

"1/23/2012 07:20:31 - Info bpbrm(pid=7212) from client xxxxxxxxx: TRV - /home/db2 is in a different file system from /home. Skipping."

The message is perfectly normal when a backup for /home is done. Note the Info status.
/home/db2 is a separate filesystem and will be backed up when it gets to that filesystem.

On one of our Solaris servers:

# ./bpmount -i
MNT - ufs 14 1 /dev/md/dsk/d0 /
MNT - local 8 8 /devices /devices
MNT - ctfs 4 16 ctfs /system/contract
MNT - PROC 4 5 proc /proc
MNT - mntfs 6 11 mnttab /etc/mnttab
MNT - tmp 4 17 swap /etc/svc/volatile
MNT - objfs 5 14 objfs /system/object
MNT - local 7 17 sharefs /etc/dfs/sharetab
MNT - fd 2 7 fd /dev/fd
MNT - tmp 4 4 swap /tmp
MNT - tmp 4 8 swap /var/run
MNT - ufs 15 9 /dev/md/dsk/d20 /one
MNT - ufs 15 4 /dev/md/dsk/d11 /two
MNT - ufs 15 12 /dev/md/dsk/d31 /three
UNAME - 5 7 4 17 5 SunOS dpurple 5.10 Generic_137137-09 sun4u
VERSION - 22 NetBackup-Solaris9 6.5
EXIT STATUS 0: the requested operation was successfully completed

The directive ALL_LOCAL_DRIVES in this instance would be like having a backup selection of:

/
/one
/two
/three

So, if I were to back up ALL_LOCAL_DRIVES on this particular server then, when NetBackup backs up the root filesystem whose 'content' obviously also includes the other ufs filesystems /one, /two and /three, I would also expect to see this 'skipping' message for all three of these different filesystems. But they would also be backed up as we are using the ALL_LOCAL_DRIVES directive.

Also as stated above, do not use cross mount points when using ALL_LOCAL_DRIVES on a *NIX client - you'll probably end up backing up more than you bargained for!

http://www.symantec.com/business/support/index?pag...

turn on multi streams

when / backs up it will say skipping anything that is a seperate file system - and / will run an its own job

then all the other filesystems will also run as their own jobs

so you will get a job for /home/db2 and for each of the others

this allows you to show it is getting the backup.

As mentioned above by couple of folks, those messages "is in a different file system from... Skipping" have been there since day-one of NBU. They have always been send from NBU Agent to NBU Server(s).

However, earlier releases of NBU had their GUI's filter out these "TRV / Trivial" messages from being actually displayed in the Activity Monitor's Job Details. The only place we would have seen such messages would have been in backup bpbkar verbose log, which is turned off by default. Later versions of NBU GUI's decided to remove the filtering, and now everything may be making its way to the GUI's Activity Monitor's Job Details.

If you really believe that you do not want those TRV/Trivial messages in the Activity Monitor's Job Details, escalate it to NBU GUI engineering for a corrective fix.

I had a same situation. I left the follow NFS and croos mount points checked and changed backup selection. instead of ALL_LOCAL_DRIVES i put all backup selection separatly. After that i got everything backed up.

I've provided the auditor with the online KB page about bpcoverage as well as the output for the specified machine.  We'll see if that is accepted or if I have to start doing strange things to keep them happy.