Roger,
I'll defer to an IBM answer, but my read is "no", they aren't recommending DISK
over FILE. I think what they are saying is no more along these lines: IFF you
are using DISK storage pools, they will perform better on Raw LV than on cooked
JFS2. This is a change from the past, where they had stated a preference for
JFS2 over RAW.
I do not think they are stating a preference for DISK over FILE.
my 2 cents.
..Paul
At 04:27 PM 8/6/2010, Roger Deschner wrote:
>I thought the food fight between sequential (devtype FILE) and random
>(devtype DISK) primary storage pools had been settled a while ago in
>favor of sequential. Indeed, TSM development has added features to make
>sequential FILE stgpools quite a bit easier to deal with.
>
>But now I am reading in "TSM 6.2 Performance Tuning Guide" that:
>"Customer experience and measurements in the lab show that raw logical
>volumes offer better client backup/restore throughput and server
>administrative process performance." (You can only allocate devtype DISK
>in raw logical volumes.) It also notes that the previous advantage of
>read-ahead provided by some filesystems such as JFS2 is now incorporated
>into many disk subsystems, so that is no longer a consideration.
>
>So it would seem that if you aren't using TSM deduplication, the
>Performance Tuning Guide now recommends that you should use random
>instead of sequential primary 1st-level disk storage pools.
>
>Am I reading this right?
>
>Roger Deschner University of Illinois at Chicago rogerd AT uic DOT edu
> Academic Computing & Communications Center
>======I have not lost my mind -- it is backed up on tape somewhere.=====
--
Paul Zarnowski Ph: 607-255-4757
Manager, Storage Services Fx: 607-255-8521
719 Rhodes Hall, Ithaca, NY 14853-3801 Em: psz1 AT cornell DOT edu
|