UNIX and Flashbackup

Yep. Create bpfis and bpbkar directories on the client under /usr/openv/netbackup/logs.

New backup attempt will write log files (log.date) to one or both of these folders.
Snapshot attempt will be logged in bpfis.

Please copy logs to process name (e.g. bpfis.txt) and upload as File attachments.

Default logging level is 0. Most errors can be picked up with level 0 logs. If not, you will have to increase logging level by adding:
VERBOSE = 5
to /usr/openv/netbackup/bp.conf

(WARNING: I am allergic to level 5 logs... If needed, someone else will have to assist with analysis...)

Please excuse the post above your, that was on the master server.  Attached is from the server in question.

Please forgive the noob, bpfis log? 

Edit....Nevermind here is the log..

Edit 2.  After looking at this log, there is no mention of the policy and server I was using (FINUX112)

Edit 3....That log had nothing to do with this problem and has been removed.

We need to find the reason for this error in bpfis log:

VxFS_Snapshot_mount: V_mount() failed with status=77, exit_status=1(Not owner)

(Will be back in a couple of hours...)

I cannot find anything related to your error. The closest that i could find was: 

VxFS_Snapshot_mount: V_mount()failed with status=77, exit_status=16(Device busy)

But it is clearly not the same.

Have you in the meantime checked permissions and ownership on snapshot device as well as device to be backed up?

Hmmm, I'm more interested in this snippet:

09:54:29.911 [9225] <2> do_cmd:  argv[6]=/dev/rdisk/isk14

Is it really configured for /isk14?  Earlier up there you said /disk14.  (Actually, Marianne also asked if "rdisk" wasn't supposed to be "rdsk" ...so maybe TWO typos?)

I reckon you might have permission problems on a disk that didn't actually exist.  :) 

Confirm the path!  See if you can find that/those typo/typos and correct it/them, and see if that helps any.

If not, we'll probably need that same log again.

I have requested that the UNIX guys take a look at the paths to answer the questions that were presented.  Thanks for all the help.

I was wondering about the same entry as Chris, but I also see this in the beginning of the log:

09:54:28.418 [9225] <2> logparams: bpfis create -nbu -owner NBU -id finux112_1355324121 -bpstart_to 300 -bpend_to 300 -backup_copy 0 -ru root -pt 20 -clnt finux112 -rg other -fim VxFS_Snapshot:cache=/dev/rdisk/disk14 -class z-test-FINUX112-Test -sched OneTime -fso -S finms-nbmaster -jobid 93768 

09:54:28.418 [9225] <2> bpfis main: received FIM as [37] VxFS_Snapshot:cache=/dev/rdisk/disk14

(Still wondering about 'rdisk' and why we see 3 different cache volumes across the thread)

So, it seems that 'do_cmd' is mis-interpreting arguments?

I would like to see results with one of the other cache volumes, e.g. /dev/rdsk/c4t0d0.

From the UNIX team

This is HP-UX 11.31, so the server has native multipathing.  You may want to post this:

some_filer:/vol/vol_images_cache_test/lun1 (LUN 1)

Agile Filename: /dev/rdisk/disk15       VG: /dev/vgimage_test

    50g (53687091200)    lun state: GOOD

Filer_CF_State: Cluster Enabled  Filer Partner: some_filer

Multipath_Policy: A/A

Multipath_Provider: HP A/A

-------- --------- -------------------- ------ ------- ------- -------------

host     filer                 /dev/dsk        primary partner HP A/A

path     path                  filename host   filer   filer   path failover

state    type          or Hardware Path HBA    port    port    priority

-------- --------- -------------------- ------ ------- ------- -------------

up       primary        /dev/dsk/c4t0d1 fcd0   0f              0

up       primary        /dev/dsk/c5t0d1 fcd0   0b              0

up       primary       /dev/dsk/c11t0d1 fcd2   0d              0

up       primary       /dev/dsk/c10t0d1 fcd2   0h              0

up       secondary      /dev/dsk/c6t0d1 fcd0           0b      1

up       secondary      /dev/dsk/c7t0d1 fcd0           0f      1

up       secondary     /dev/dsk/c12t0d1 fcd2           0h      1

up       secondary     /dev/dsk/c13t0d1 fcd2           0d      1

 

some_filer:/vol/vol_cache_test/lun0 (LUN 0)

Agile Filename: /dev/rdisk/disk14

    10g (10737418240)    lun state: GOOD

Filer_CF_State: Cluster Enabled  Filer Partner: some_filer

Multipath_Policy: None

Multipath_Provider: None

-------- --------- -------------------- ------ ------- --------

host     filer                                 primary partner

path     path                  /dev/dsk host   filer   filer

state    type                  filename HBA    port    port

-------- --------- -------------------- ------ ------- --------

up       primary        /dev/dsk/c4t0d0 fcd0   0f

up       primary        /dev/dsk/c5t0d0 fcd0   0b

up       secondary      /dev/dsk/c6t0d0 fcd0           0b

up       secondary      /dev/dsk/c7t0d0 fcd0           0f

up       primary       /dev/dsk/c11t0d0 fcd2   0d

up       primary       /dev/dsk/c10t0d0 fcd2   0h

up       secondary     /dev/dsk/c12t0d0 fcd2           0h

up       secondary     /dev/dsk/c13t0d0 fcd2           0d

 

The Agile Filename is the "disk" device that HP-UX creates so that when you use storage it can failover in the background.

I wonder if these multi-pathed disks are supported?

Maybe Chris can help here:

In Snapshot Client Admin Guide I see the following: