We've toyed with using a disk pool based on devclass devtype=file. The
upside is this pre-allocation problem is not an issue. The downside is the
definition for the device class doesn't allow you to specify multiple
devices. So there needs to be some way to aggregate your devices into one
big thing (and RAID5 is probably NOT the right thing to do).
There are other issues with this approach, like reclamation and caching,
etc., that should be considered. But this idea might have merit in this
case.
Kelly J. Lipp
STORServer, Inc.
485-B Elkton Drive
Colorado Springs, CO 80907
lipp AT storsol DOT com or kelly.lipp AT storserver DOT com
www.storsol.com or www.storserver.com
(719)531-5926
Fax: (240)539-7175
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU]On Behalf Of
Matt Simpson
Sent: Wednesday, October 23, 2002 5:51 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Co-location
At 12:06 AM -0500 10/23/02, Chris Gibes said:
>You are absolutely correct. Co-location is by storage pool, not by
>management class. So yes, you would need to carve your disk up into
>multiple storage pools to selectively use co-location,
Thanks for the confirmation
>my guess is that it's
>not viable to add more disk (or you're on one of those platforms where
>disk is not "cheap"...)
Expenditures are always political. Management is always more willing
to spend huge gobs of money into a new disaster than drop a few more
pennies into an existing one. And I'm more concerned about the
management headaches than the cost, as you point out ..
>the total amount of disk and the total amount
>being backed up are going to be the same regardless of how many pools
>you have, so carving one big pool up, shouldn't be that big of an issue,
true, but
>as long as you put some planning into the size of the disk pools.
There's the catch. We can't plan more than 30 minutes into the
future around here. It's easier to manage one big chunk of something
than a bunch of little chunks. If we carve up our disk pools based
on today's "plan", we'll have to re-configure them tomorrow. Our
database has already exceeded the allocation that IBM told us was way
bigger than we'd ever need, and we haven't even finished the
installation yet.
--
Matt Simpson -- OS/390 Support
219 McVey Hall -- (859) 257-2900 x300
University Of Kentucky, Lexington, KY 40506
<mailto:msimpson AT uky DOT edu>
mainframe -- An obsolete device still used by thousands of obsolete
companies serving billions of obsolete customers and making huge obsolete
profits for their obsolete shareholders. And this year's run twice as fast
as last year's.
|
