Hi Becky,
We run STK 9710 silos attached to AIX hosts.
I'm sure you will get several different opinions and points of view on these
questions, but here are mine:
1) The 9710 robot microcode is provided by STK. The DLT 7000 drive
microcode comes from Quantum, but will be installed on the drives by your
STK CE's.
The microcode level needed for the robot depends on your setup, so rely on
what your CE tells you. Make sure the DLT7000 microcode is at least up to
V7.
2) I find that tape operations are easier in one device class, one library.
If you check your tapes in to the library as scratch, then when it's time to
do a DB backup, ADSM just grabs the next scratch tape; and when the DB
backup expires, it just goes back to the scratch pool. (just the same as
using 1 scratch pool with MVS TMS). If you split the library, then you have
to constantly check to make sure you have enough scratch tapes in each
category. If everything is in one devclass/scratch pool, it makes your life
easier. If you are worried about one storage pool eating too many scratch
tapes, you can control that with the MAXSCRATCH parmeter on the storage pool
itself.
3) Some people manage their offsite/onsite movement via unix scripts. You
have to identify which tapes you want to move offsite, check them out of the
ADSM library, eject them from the robot, and update the status of each moved
volume to show it is no longer available to mount.
We use DRM instead. I am familiar with TMS - what DRM provides you is the
function that is equivalent to TMS vaulting. With DRM, any tapes created in
DRM-eligible COPY pools are automatically flagged as eligible for movement
offsite; then DRM does the data base updates for you as tapes move from
onsite to your courier to your vault and back.
Unless you have extra humans sitting around with nothing to do but write and
maintain scripts, I think DRM pays for itself.
Hope this helps.
===============================================================
Wanda Prather
Johns Hopkins Applied Physics Lab
443-778-8769
wanda_prather AT jhuapl DOT edu
"Intelligence has much less practical application than you'd think."
- Scott Adams/Dilbert
===============================================================
> ----------
> From: Buechler, Becky[SMTP:buechler AT kcc DOT com]
> Sent: Wednesday, July 22, 1998 11:16 AM
> To: ADSM-L AT vm.marist DOT edu
> Subject: STK 9710 question
>
> I have been working with ADSM server's on MVS for the past two years.
> Now I will be adding another ADSM server on a HPUX platform using the
> STK 9710 silo. I am new to the unix environment and tape devices and
> have a few questions I was hoping someone could help me with. The
> drives are DLT7000.
>
> 1. What microcode level does the 9710 silo and drives need to be at? I
> remember seeing some information on this list before pertaining to
> problems if you aren't at a certain level.
>
> 2. Is it best to add all the carts to ADSM under one device class or
> is it better to add multiple ones. Maybe one deviceclass for database
> backups, one for copystorage pool and one for the onsite storage pool?
>
> 3.) With the MVS platform we have TMS for handling carts. Everything
> is pretty automatic for sending carts offsite. Is their anything like
> this in the unix environment? If not, are the lists created using
> scripts? Or do most people use DRM?
>
> I would appreciate any comments or advice pertaining to the STK 9710
> silo.
>
> TIA
>
> Becky Buechler
> Kimberly Clark Corporation
> email: buechler AT kcc DOT com
>
|
