topic Re: SYSMAN IO AUTOCONFIGURE Question in Operating System - OpenVMS

SYSMAN IO AUTOCONFIGURE Question

Phillip Tusa — Thu, 17 Jan 2008 15:22:44 GMT

Greetings to all ...

I am considering running SYSMAN IO AUTOCONFIGURE on an itanium rx-2620. I am considering doing this because when I try to run a full image backup, I see the following info messages:

%MOUNT-I-OPRQST, medium is offline
%MOUNT-I-OPRQST, Please mount device _$2$MGA2:

At this point the process appears to "hang" and cannot go any futher.

I also get the same info messages when I attempt to issue the command:

$MOUNT /FOREIGN $2$MGA2: /OVER=ID

The Tape Library is a MSL-5000. We have three itaniums clustered together. We have OPEN VMS version 8.2-1 loaded on each node.

I had no problem about 6 months ago performing a full image backup.

Has anyone else had a mag tape problem like this? Any assistance with this issue would be greatly appreciated!

Thank you;

Phil

Re: SYSMAN IO AUTOCONFIGURE Question

Steven Schweda — Thu, 17 Jan 2008 16:06:01 GMT

> [...] SYSMAN IO AUTOCONFIGURE [...]

This will create a device, if it's needed.
You already have the device, or you'd be
getting:

%MOUNT-F-NOSUCHDEV, no such device available

for which symptom, SYSMAN IO AUTOCONFIGURE
might actually help.

> At this point the process appears to "hang"
> and cannot go any futher.

If you don't have an operator waiting to
service these requests, you might consider
adding /NOASSIST to commands like this, but
that won't solve the basic problem, which
appears to be an unhappy tape drive. What
makes you think that the tape drive is ready
to use?

Re: SYSMAN IO AUTOCONFIGURE Question

Phillip Tusa — Thu, 17 Jan 2008 16:27:54 GMT

> What makes you think that the tape drive is ready to use?

OK .. when I do this:
$SH DEV MGA

The resultant message shows the devices
(i.e. $2$MGA2: or $2$MGA3:) in an online status.

Re: SYSMAN IO AUTOCONFIGURE Question

Hoff — Thu, 17 Jan 2008 16:32:42 GMT

And to continue the line of investigation suggested and started by Steven, start looking at the Fibre Channel configuration, and do look at using an OpenVMS I64 V8.3 or later DVD for your testing.

Or do build a bootable OpenVMS I64 system disk in an available disk in your existing configuration, and use that.

When you performed the full BACKUP /IMAGE six months ago, was it the same command and the same target in the same configuration? Or is the use of the FC SAN tape new? Or has the FC SAN been reconfigured? (Yes, this is the "what might have changed?" question.)

There are various steps required around configuring FC SAN devices. (The use of $2$ and MG tends to indicate this is an FC SAN tape.) If the device is accessible from your running OpenVMS environment, then the issues would appear local to the (I assume) OpenVMS bootable DVD environment; to the install disk and its configuration.

FWIW, the medium is off-line indication in the MOUNT diagnostics indicates that the MGDRIVER device driver or the MSL5000 device is unhappy about the status of the device, or with the status of the medium loaded within the device.

The MSL5000 tape library is usually run sequentially through the media present in the path, or it is controlled with the MRU tool. IIRC, if the drive is set for ROBOT, you likely want to use MRU.

Re: SYSMAN IO AUTOCONFIGURE Question

Steven Schweda — Thu, 17 Jan 2008 16:42:33 GMT

> OK .. when I do this:
> $SH DEV MGA
>
> The resultant message shows the devices
> (i.e. $2$MGA2: or $2$MGA3:) in an online
> status.

Not conclusive (by a long shot). On my
Alpha, for example:

alp $ show devi /full dlt

Magtape ALP$MKB400:, device type Quantum DLT4000 CPQ DRV, is online, record-
oriented device, file-oriented device, served to cluster via TMSCP Server,
error logging is enabled, controller supports compaction (compaction
nabled), device supports fastskip (per_io).

Error count 4 Operations completed 2615649
Owner process "" Owner UIC [SYSTEM]
Owner process ID 00000000 Dev Prot S:RWPL,O:RWPL,G:R,W:RWPL
Reference count 0 Default buffer size 512
Density TK87 Format Normal-11

Volume status: no-unload on dismount, beginning-of-tape, odd parity.

This looks very impressive, but the power to
the (external) drive is off, so if I tried to
use it, it would be a miracle if it worked.

A better answer would be something like,
"There's a tape of the right type in the
right place, and all the lights on the drive
say that it's ready to go."

Re: SYSMAN IO AUTOCONFIGURE Question

The Brit — Thu, 17 Jan 2008 17:12:54 GMT

Can you confirm that you have done a

$ Robot Load Drive X Slot Y $1$GGAZ

(where the tape behind the drive X is $2$mga2:)

You will get this message if the tapedrive has not been loaded.

Dave

Re: SYSMAN IO AUTOCONFIGURE Question

DECxchange — Fri, 18 Jan 2008 01:32:26 GMT

Hello,
I don't know much about SANs, but for any other type of device "medium is offline" message, it could be:
cable unplugged or not securely plugged into device
power is not on
tape or CD is not in the drive.
tape is stuck in the drive.
The drive door is open.
Error light is flashing
hardware failure

Or something silly like that.

Re: SYSMAN IO AUTOCONFIGURE Question

Mark Caccavone_1 — Fri, 18 Jan 2008 14:47:16 GMT

This is because you have no tape mounted in the drive.

You should have a NSR attached which you should see as a $2$GGA0:

To see if the tape is mounted in the drive enter :-

$ROBOT SHOW ROBOT $2$GGA0 DRIVES

If drive 2 shows as empty then you need to load a tape from the MSL into the drive.

e.g.

$ROBOT LOAD ROBOT $2$GGA0 DRIVE 2 SLOT 1

This will take a tape from slot 1 in the library and load it into $2$MGA2:

If you cannot see $2$GGA0 , then you may need to do a IO AUTO withing SYSMAN. If you cannot see it after that - then you are ZONED out of seeing the NSR. Which gets a bit more complicated.

Re: SYSMAN IO AUTOCONFIGURE Question

Anton van Ruitenbeek — Thu, 07 Feb 2008 15:00:38 GMT

Phil,

We had simular problems using an old library acros a NSR. But I asume you have connected the MSL directly via a fibre-switch.
The problem whe never know the answer for except rebooting the NSR. I still think there was a problem in the SCSI cable used to the library. I'm not anymore whith that costumer so I can't reproduce this problem.

Do you have this problem on all nodes or is the MSL connected to only one I64 ?

AvR