Confused NBU Environment

[Marianne]

Please explain what you are confused about?

Or is NetBackup 'confused'?

Are you seeing any errors?

NBU version? Os version?

[grosstech]

Hi Marianne:

    I was starting to detail the environment and got disconnected from the Internet.  So,

not only is NBU confused but so am I!!

Environment:

NBU Master: Sun Ultra 4500 8 CPU old but .. Solaris 9 w/ NBU 5.1

NBU Media Server: Sun Ultra 420R (Ultra 80 Chassis) Solaris 9 w/NBU 5.1

Enterprise Tape Library: ADIC Scalar 100 w/ 6 IBM Ultrium LTO-2s in shared configuration.

I believe that there is also a StorageTek array of disks connected which hold the Veritas catalog.

History:  Approximately 2 weeks ago the ADIC began dropping and mangling tapes.  Third party vendor supporting the unit worked on it for several days and could not repair the picker.  They flew a second "new" unit in and pulled the "Data Router" what I would term as the FC electronics from the first unit and 5 of the 6 tape drives.  One drive in the old unit had broken.  System worked for a few days and then dropped tapes.  Worked again until last Wednesday when it dropped tapes.  A reboot of the library and I did not see any devices from Veritas.  Turned out that the ADIC had not properly presented itself to the switch.  Hard boot and it seems to be presenting but no backups are running.

cfgadm -al reveals:

c5 fc-fabric connected configured unknown

c5::200200308c005a0e array-ctrl connected configured unknown

c5::200500a0b82134f9 disk connected configured unknown

c5::200500a0b82134fa disk connected configured unknown

c5::2100001b32055019 unknown connected unconfigured unknown

c5::2100001b320662d1 unknown connected unconfigured unknown

c5::2100001b320f56aa unknown connected unconfigured unknown

c5::2100001b320f5c8f unknown connected unconfigured unknown

c5::210000e08b11a286 unknown connected unconfigured unknown

c5::210000e08b9ceef7 unknown connected unconfigured unknown

c6 fc-fabric connected configured unknown

c6::200100308c005a0e array-ctrl connected configured unknown

c6::200400a0b82134f9 disk connected configured unknown

c6::200400a0b82134fa disk connected configured unknown

c6::2100001b3205ff19 unknown connected unconfigured unknown

c6::2100001b320652d1 unknown connected unconfigured unknown

c6::2100001b320903ee unknown connected unconfigured unknown

c6::2100001b320fd1ac unknown connected unconfigured unknown

c6::210000e08b116d86 unknown connected unconfigured unknown

c6::210000e08b9ca6f7 unknown connected unconfigured unknown

sgscan shows from the Master:

/dev/sg/c0t10l0: Disk (/dev/rdsk/c0t10d0): "SEAGATE ST336605LSUN36G"

/dev/sg/c0t11l0: Disk (/dev/rdsk/c0t11d0): "SEAGATE ST336605LSUN36G"

/dev/sg/c0t6l0: Cdrom: "TOSHIBA XM6201TASUN32XCD"

/dev/sg/c0tw200100308c005a0el1: Changer: "ADIC Scalar 100"

/dev/sg/c0tw200100308c005a0el2: Tape (/dev/rmt/5): "IBM ULTRIUM-TD2"

/dev/sg/c0tw200100308c005a0el3: Tape (/dev/rmt/4): "IBM ULTRIUM-TD2"

/dev/sg/c0tw200200308c005a0el4: Tape (/dev/rmt/3): "IBM ULTRIUM-TD2"

/dev/sg/c0tw200200308c005a0el5: Tape (/dev/rmt/2): "IBM ULTRIUM-TD2"

/dev/sg/c0tw200200308c005a0el6: Tape (/dev/rmt/1): "IBM ULTRIUM-TD2"

/dev/sg/c0tw200200308c005a0el8: Tape (/dev/rmt/0): "IBM ULTRIUM-TD2"

/dev/sg/c0tw200400a0b82134f9l0: Disk (???): "STK OPENstorage D280"

/dev/sg/c0tw200400a0b82134fal0: Disk (???): "STK OPENstorage D280"

/dev/sg/c0tw200500a0b82134f9l0: Disk (???): "STK OPENstorage D280"

/dev/sg/c0tw200500a0b82134fal0: Disk (???): "STK OPENstorage D280"

From the Media Server:

/dev/sg/c0t0l0: Disk (/dev/rdsk/c0t0d0): "FUJITSU MAN3367M SUN36G"

/dev/sg/c0t1l0: Disk (/dev/rdsk/c0t1d0): "FUJITSU MAN3367M SUN36G"

/dev/sg/c0t6l0: Cdrom: "TOSHIBA DVD-ROM SD-M1401"

/dev/sg/c2t4l0: Array-controller: "ADIC SNC 6101"

/dev/sg/c2t4l4: Tape (/dev/rmt/0): "IBM ULTRIUM-TD2"

/dev/sg/c2t4l5: Tape (/dev/rmt/1): "IBM ULTRIUM-TD2"

/dev/sg/c2t4l6: Tape (/dev/rmt/2): "IBM ULTRIUM-TD2"

/dev/sg/c2t4l8: Tape (/dev/rmt/3): "IBM ULTRIUM-TD2"

/dev/sg/c3t4l0: Array-controller: "ADIC SNC 6101"

/dev/sg/c3t4l1: Changer: "ADIC Scalar 100"

/dev/sg/c3t4l2: Tape (/dev/rmt/4): "IBM ULTRIUM-TD2"

/dev/sg/c3t4l3: Tape (/dev/rmt/5): "IBM ULTRIUM-TD2"

From the master by manipulating/guessing at the parameters for -mt and -ml I am able to sometimes get one (not two) Array-controller:"ADIC SNC 6101" into the sgscan.  Currently it is not there.  I have also sometimes only been able to get this with a reboot with a reconfigure.

tpconfig -dl from the Master:

Currently defined drives and robots are:

Drive Name IBMULTRIUM-TD20

Index 0

NonRewindDrivePath /dev/rmt/2cbn

Type hcart2

Status DOWN

Shared Access Yes

TLD(0) Definition DRIVE=1

Drive Name IBMULTRIUM-TD21

Index 1

NonRewindDrivePath /dev/rmt/4cbn

Type hcart2

Status DOWN

Shared Access Yes

TLD(0) Definition DRIVE=2

Drive Name IBMULTRIUM-TD22

Index 2

 NonRewindDrivePath /dev/rmt/3cbn

Type hcart2

Status UP

Shared Access Yes

TLD(0) Definition DRIVE=3

Drive Name IBMULTRIUM-TD23

Index 3

NonRewindDrivePath /dev/rmt/5cbn

Type hcart2

Status DOWN

Shared Access Yes

TLD(0) Definition DRIVE=4

Drive Name IBMULTRIUM-TD24

Index 4

NonRewindDrivePath /dev/rmt/1cbn

Type hcart2

Status DOWN

Shared Access Yes

TLD(0) Definition DRIVE=5

Drive Name IBMULTRIUM-TD25

Index 5

NonRewindDrivePath /dev/rmt/0cbn

Type hcart2

Status DOWN

Shared Access Yes

TLD(0) Definition DRIVE=6

Currently defined robotics are: TLD(0) robotic path = /dev/sg/c0tw200100308c005a0el1,

volume database host =

When I go into the GUI and ask it to configure Storage Devices it comes back with One Robot for the Master and 8 drives.  I believe that perhaps the bad drive is still there but not sure how to identify or get rid of.  On the Media Server it says that there is one Robot but 12 drives.  Is the second array controller causing this sort of output.  I would expect for proper configuration I need one Robot and 6 drives configured.  None of the backup jobs when started progress beyond a mount and connection.  The tapes are generally frozen and jobs appear active which don't run and others are queued and eventually fail.  So, I believe that both NBU and myself are confused.  Any help will be greatly appreciated.  I can provide additional details if desired.  Thanks.

Mark

[grosstech]

Hi Marianne:

    I'm new to the forum.  I hope that posted everything that was necessary.  Not sure if you saw it.  Thanks for any help.

Mark

[mph999]

Forget NBU, your issues are at a hardware / os level.  Trying to fix this using NBU info will not help, you must work at the os level.

I suspect on your system cfgadm should you tape devices, as opposed to unknown.

You could download mtx (for solaris) available I think from http://sunfreeware.com/

I have a copy on my laptop which I can send to you if ou send me an email :

martin<underscore>holt@symantec.com

I also have a brief 'howto' guide to get it working.  Please note this is not Symantec software, and is therefore 'unsupported'.

In a nutshell, it is siilar to mt but allows control of the robot outside NBU, useful for troubleshooting/ testing.

First thing I would do is powercycle everything, server, switch, drives, robot - not reset, off/ on.

THen, if no luck, see what's visible at a switch level - I have seen cfgadm show all devices, even though NBU refused to work with them, turns out there was a san issue (unknow deteails, sorry), so although we could see them, we couldn't talk properly - reason I mention this is just to 'show' that anythings possible ...

If after the powercycle things are not up and running, you may need to talk to the library vendor/ hardware support again, as I suspect this is where the issue is.

Only when cfgadm and mtx are fully wroking, should you consider revisiting NBU.

Regards,

Martin (Sym NBU TSE, UK)