RE: BRMS RSTLIBBRM jobs impacting RDI and LPAR performance

Rob,

I've already have this.

5 Network Storage Servers with 6 Network drives each, 100gb each, (Total of 3 TB) hosted from Production SSD to client R&D ins ASP02.
Two of our 7 R&D environments are 100 % NWS SSD.
Base OS is all 10k spinny.

Performance on these two NWS SSD environments is 4x that of the 10k spinny environments, matching Production SSD run times.

Paul

-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Rob Berendt
Sent: Tuesday, July 03, 2018 3:31 PM
To: Midrange Systems Technical Discussion
Subject: RE: BRMS RSTLIBBRM jobs impacting RDI and LPAR performance

Here is a radical idea. Unload/reload the whole enchilada. Instead of
just a Production lpar and a Development lpar have three:
Host, Production, Development.
Host would own all the disk, SSD and spinning. Put SSD and spinning into
two different ASP's. I've done this on three machines. Two were to break
up older/vs newer SSD's. One was to break up faster spinning disks from
larger slower archival spinning disks.
This allows you to put load/source on all partitions on the drives of your
choice, like maybe all load/source on SSD? And you can break up critical
libraries from not so critical libraries on to disk type also.
Now, I've not played with separate ASPs on the guests but that can
probably be done.
Here I'm talking about the older "traditional" ASPs and not IASPs.
Also you can add disk from the fly from the host on to the guests. I do
that.

CRTNWSSTG NWSSTG(LPARA101)
NWSSIZE(35000)
FORMAT(*OPEN)
ASP(2)
TEXT('LPAR A, NWSD 1, Disk 01')



Rob Berendt