Oooh, I like that idea. I had no idea that the LTO2 drives
worked that way. Indeed if the 3592s work that way, I would love the WWN
to stay the same when the drives are swapped out.
Now, if I could only find some documentation to tell me if it
indeed worked that way....
Thanks,
Ben
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Kauffman, Tom
Sent: Wednesday, October 27, 2004 3:46 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Adding 3592's to a 3494
Ben --
No *real* idea on the 3592 hard/soft addressing -- but the LTO2 drives
in a 3584 use a synthetic wwn based on 3584 serial number and drive
position (frame and slot) so the wwn doesn't change if the drive gets
swapped.
So -- my guess would be that 'hard' addressing would be the drive's
native wwn and 'soft' addressing would be a synthetic wwn based on s/n,
frame, and slot.
This may have something to do with your cfgmgr issue.
Tom Kauffman
NIBCO, Inc
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Ben Bullock
Sent: Wednesday, October 27, 2004 3:56 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Adding 3592's to a 3494
Matthew (and others in the group),
Just this week we put in our 2 test 3592 drives and started
playing with them. These are sweet. So far they are running a little
over 2X the speed of the 3590E drives, and we have yet to test the
capacity.
We've been able to test out the alternate pathing, dynamic load
balancing, etc.. It's all through a SAN, so we have tested the zoning,
etc. Failovers work as expected with no interruption to the TSM server
at all. Very nice.
AIX 5.2, TSM 5.2.1.3, Atape 9.0.7.0, atldd 5.5.1.0, 3592 drives,
3490 library.
All has gone smoothly until we went to simulate a failed drive
where we would swap it out for a new one. It didn't go well, with TSM or
the OS not able to see the new drive correctly. When I tried to 'cfgmgr'
it back in, it gave me these errors:
root:># cfgmgr
Method error (/usr/lib/methods/cfgppa_isa -l ppa0 ):
0514-038 Error loading kernel extension.
Method error (/etc/methods/cfgtsmdd -l mt0 ):
0514-051 Device to be configured does not match the physical
device at the specified connection location. Method
error (/etc/methods/cfgtsmdd -l mt1 ):
0514-051 Device to be configured does not match the physical
device at the specified connection location.
I had to remove these 2 "mt" devices to get it to stop erroring.
That's a new one to me...
I have 2 questions:
1: During the installation of the tape drives the IBM CE asked
if we wanted "hard" or "soft" addressing. His instructions say to do
HARD, but then his support folks said "soft", what should it be? I'll be
calling this in to IBM, but was curious if anybody here had an answer.
2: On our switch we are zoning the drives and FC adapters by
WWN. Doing it this way, am I going to have to totally remove and
recreate a drive when I swap it out? (i.e. remove from TSM, the OS,
unzone the switch. Replace drive, re-configure it all back in.).
Thanks,
Ben
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Matthew Glanville
Sent: Thursday, September 30, 2004 11:34 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Adding 3592's to a 3494
When we upgraded we needed a new library defined in TSM to use the
3592's with different private/scratch categories, and device class's,
drives, etc.. Not a different 'logical' library defined on the 3494
library manager. That may have worked too, but it depends on how
particular you want to be as to which drives/slots are available to each
host connected to the library.
3592's have been good. So far after 9 months use of 500 tapes and 6
drives.
2 tapes had I/O problems, no serious problem in data lost, TSM's
'restore volume' cleaned it up in minutes. some drives have just needed
code upgrades to remove minor problems. One had 'clean me' displayed
permanantly, it went way after that. One was replaced, but it was bad
from the factory on initial installation.
Try to do the code upgrades for the drives from the Host's fiber/scsi.
When the IBM SE pushed it via the Library manager it took hours instead
of minutes.
The web interface on the upgraded library manager is great!
No longer do you have to login to the library manager console to find
out why 'operator intervention' is required... You can check it from
your desk and decide to call in IBM to fix the problem or just unload
the tapes that wanted to come out of the I/O slots.
My biggest issue is that the server we have connected to those 6 drives
can't push data fast enough to them, they can go much faster!
Matthew Glanville
Eastman Kodak
|
