Rob,
You need to get after the work management a bit. I'd start with putting the
network servers into a shared pool. You will have to vary them off and on
again which of course means shutting down your partition. If the
performance of the hosted partition is OK then it will wait until you have a
maintenance window.
Unless, you've got a situation (and guts enough) where you can vary off one
server, make the change and vary it back on again quickly. You'll get a
really nasty SRC about losing devices (disk in this case) and then when it
varies back on it will see the disk units again and happily take off again.
Give the partition time to get its collective stuff together then do the
next one. Ugly, but it'll work if you simply cannot IPL the partition.
Don't ask how I know it will work.........
--
Jim Oberholtzer
Chief Technical Architect
Agile Technology Architects
-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
rob@xxxxxxxxx
Sent: Tuesday, October 21, 2014 5:52 AM
To: Midrange Systems Technical Discussion
Subject: RE: Performance considerations of a disk hosting lpar Was: Disk:
Percent full a performance factor?
WRKNWSD
Work with Network Server Descriptions
5=Display
Pool identifier . . . . . . . . . : *BASE
System pool identifier . . . . . : 2
WRKSHRPOOL
Defined Max Allocated Pool -Paging Option--
Pool Size (M) Active Size (M) ID Defined Current
*MACHINE 947.58 +++++ 947.58 1 *FIXED *FIXED
*BASE 6278.33 145 6278.33 2 *FIXED *FIXED
*INTERACT 811.89 203 811.89 3 *FIXED *FIXED
*SPOOL 81.18 5 81.18 4 *FIXED *FIXED
*SHRPOOL1 .00 0 *FIXED
*SHRPOOL2 .00 0 *FIXED
*SHRPOOL3 .00 0 *FIXED
*SHRPOOL4 .00 0 *FIXED
*SHRPOOL5 .00 0 *FIXED
Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1 Group Dekko Dept 1600 Mail
to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com
From: "Jim Oberholtzer" <midrangel@xxxxxxxxxxxxxxxxx>
To: "'Midrange Systems Technical Discussion'"
<midrange-l@xxxxxxxxxxxx>
Date: 10/20/2014 04:14 PM
Subject: RE: Performance considerations of a disk hosting lpar Was:
Disk: Percent full a performance factor?
Sent by: "MIDRANGE-L" <midrange-l-bounces@xxxxxxxxxxxx>
Do you have the network servers using a pool identifier (I always use
*SHRPOOL50) to push the I/O memory there? That keeps it out of *BASE.
With 10TB of storage being hosted I have 4096 GB in that pool and that's
almost overkill. You might get away with 3Gb instead.
--
Jim Oberholtzer
Chief Technical Architect
Agile Technology Architects
-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
rob@xxxxxxxxx
Sent: Monday, October 20, 2014 2:46 PM
To: Midrange Systems Technical Discussion
Subject: Performance considerations of a disk hosting lpar Was: Disk:
Percent full a performance factor?
On a loosely related note, what performance considerations should one
really
watch for on an lpar that does nothing but host disk? For example on our
Power 8's all disk (except to the vios lpars) is hosted by one lpar of IBM
i. This then hosts several lpars of IBM i and AIX.
RACK1HST: Hosts main production lpars
% CPU used . . . . . . . : 66.1
Auxiliary storage:
System ASP . . . . . . : 23277 G
% system ASP used . . : 54.3132
System Pool Reserved Max -----DB----- ---Non-DB---
Pool Size (M) Size (M) Active Fault Pages Fault Pages
1 896.18 369.86 +++++ .0 .0 1.9 1.9
2 6329.73 14.27 145 .2 1.4 6.3 30757
3 811.89 .00 203 .0 .0 .6 2.3
4 81.18 .00 5 .0 .0 .0 .0
Work with Disk Status
Elapsed time: 00:00:37
%
Busy
0
0
0
...
For all 32 SSD units.
Virtual Processors Processing units Memory
(GB)
Minimum Assigned Maximum Minimum Assigned Maximum
Minimum Assigned Maximum
RACK1HST 1 10 20 0.1 0.5 1 4 8
16
RACK2HST: Hosts Mimix lpars and backup Domino lpars.
% CPU used . . . . . . . : 5.1
Auxiliary storage:
System ASP . . . . . . : 22500 G
% system ASP used . . : 64.2344
System Pool Reserved Max -----DB----- ---Non-DB---
Pool Size (M) Size (M) Active Fault Pages Fault Pages
1 781.05 362.50 +++++ .0 .0 3.6 3.6
2 6444.86 14.21 175 .0 .0 5.8 255.6
3 811.89 .00 203 .0 .0 1.5 6.4
4 81.18 .00 5 .0 .0 .0 .0
Work with Disk Status
Elapsed time: 00:00:39
%
Busy
0
0
0
...
For all 30 SSD units.
Virtual Processors Processing units Memory
(GB)
Minimum Assigned Maximum Minimum Assigned Maximum
Minimum Assigned Maximum
RACK2HST 1 10 20 0.1 0.5 1 4 8
12
RACK3HST: Hosts rarely used test lpar and a triple backup of our Domino
lpars.
% CPU used . . . . . . . : 19.7
Auxiliary storage:
System ASP . . . . . . : 5956 G
% system ASP used . . : 45.8444
System Pool Reserved Max -----DB----- ---Non-DB---
Pool Size (M) Size (M) Active Fault Pages Fault Pages
1 691.21 352.51 +++++ .0 .0 .0 .0
2 6534.69 13.31 175 .0 .0 2.0 286.7
3 811.89 .00 203 .0 .0 1.2 8.1
4 81.18 .00 5 .0 .0 .0 .0
Work with Disk Status
Elapsed time: 00:00:29
%
Busy
0
0
0
...
for all 22 spinning disks.
Virtual Processors Processing units Memory
(GB)
Minimum Assigned Maximum Minimum Assigned Maximum
Minimum Assigned Maximum
RACK3HST 1 10 20 0.1 0.5 1 6 8
12
Rob Berendt
midrange.com
