No tape technology is perfect. They are mechanical devices. Tapes fail
(wear, stretch-and-tear), drives break. Operators drop tapes. Buildings
burn down.
So no, I wouldn't put any stock in this.
As Richard said, you want offsite, DR recovery tapes, don't you ? If an
onsite, primary tape gets eaten by the drive, you do want to be able to
recover it, don't you.
Granted, the offsite don't need to be a 1-1 match. My onsite primaries are
moving to mostly 3592-E05 and the offsites are the piece-o-!@#%^&* LTO2
drives (that we tried to use as primaries but still fail, almost weekly -
IBM admits they weren't designed for the load we put through them), since
we didn't want to throw away the $$$$$$$$ money invested in 2-3583
libraries and 1000 tapes.
Keith Arbogast <warbogas AT INDIANA DOT EDU>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
02/02/2007 11:22 AM
Please respond to
"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
To
ADSM-L AT VM.MARIST DOT EDU
cc
Subject
[ADSM-L] Writing to copy pools on 3584?
We are starting the process of replacing our 3494 tape library. It
has ten 3590-E1A drives, about 2,500 cartridges, and supports two TSM
servers.
We have had preliminary talks with IBM about the 3584 ATL and 3592-
E05 drive technologies. A technical rep told us, more or less, "No
one is making copy pools with this new technology. They do not copy
tapes from primary storage pools to a copy pool." I would never
consider this with our current tape technology, so it will be a giant
leap of faith to stop creating them no matter how good the new
technology is. However, people of my generation who think objects
that look like bombs might really be bombs are heaped with ridicule
by those who know they are merely cartoon characters.
Is this truly what most organizations are doing who use 3584 ATL's
and 3592-E05 drives? We would be grateful to hear what others are
doing, and their reasons for stopping or not stopping writing to copy
pools?
Thank you,
Keith Arbogast
Indiana University
|
