Re: IRD and Group Capacity (was: z196 tools question)

[Al Sherkow]

Kees --

IRD Weight Management is used at some of my client sites. I think it is very 
useful. For a review that capabilities this feature enables to WLM to shift 
"weight" between LPARs in the same SYSPLEX on a single CPC based on importance. 
For example if importance=3 work in LPAR A is exceeding it's goal, and 
importance=2 work in LPAR B is missing it's goal then WLM will move some 
"weight" from LPAR A to LPAR B to help out the importance=2 work.

Seems very useful to me, but I'm a consulting and I'm not running a data 
center. 

On to LPAR Group Capacity Limit. Another nice feature, requested by customers 
for a long time and introduced with z/OS 1.8. LPAR GCL is also "managed" by 
WLM. This capability manages the 4 hour rolling average of a group of LPARs to 
not exceed a customer set limit. This capability uses the LPAR's weight 
settings as the indication of how the LPARs should participate in the group. 

However as Kees wrote in the original post:

The docs mention "As WLM is responsible for
both WLM LPAR Weight Management and 
soft-capping, it will be aware when an LP is 
being soft-capped, and will not make any 
further adjustments to the weight of the LP 
until the soft-cap is removed." 

This is a problem, it is when the GCL is being enforced that a site would most 
like IRD LPAR Weight Management to kick in and start redistributing the limited 
CPU resource based on importance of the workloads rather than the initially set 
weights of the LPARs. 

IBM does not want to fix this issue. I think it has been entered as a "SHARE 
requirement".
 
So some sites have decided not to use GCL because they view "IRD LPAR Weight 
Management" as more useful. Other sites have made the decision the other way. 
Still other sites are using both while understanding this limitation. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: Are LPAR names unique or effectively unique?

[Al Sherkow]

LPAR Names are unique on a single serial number, but I have at least one client 
that has the same LPAR names on different physical machines.  


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Mainframe Software Pricing,
WLC, LPARs and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062  
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: Actually it is more typically 2% (Was Re: IBM announces 6% price increase for z/OS)

[Al Sherkow]

Sorry if my posts yesterday were confusing. I posted as a reply to the 6% post, 
but realized many would not open that old thread and that I needed to have 2% 
as part of the subject. So I "replied" again, and did not repeat the content. 
So here is the content and below that I'll respond to the questions and 
comments I've received: 

--
I've done further analysis, now that the LCS_PRICING Interactive Report is 
updated with the new prices. When viewed not tier by tier, but with a stack of 
other software products (for which the prices did not change) the increase on 
the MLC invoice due to the z/OS increase is closer to 2%. As more and more 
products are added to the software stack, the percent increase will be smaller, 
as the other products are added to both the before and after totals equally. 
One customer has shared their analysis with me and for their products and 
features the monthly impact is approximately 1.8%. 

My analysis is available at: 

http://www.sherkow.com/updates/zos_2012_price_increase.html 

I hope you find this interesting and I welcome any comments or feedback you may 
have. I can find no other public analysis that shows this effective increase is 
near 2%. 

The bottom line is that everyone needs to analyze their own site. Similar to 
determining which software pricing metrics your site should be using there 
still are not clear rule of thumbs that work for every site. 
--

My analysis was with IBM software only, and a very small, completely 
sub-capacity product set. Adding other products, whether from IBM or from an 
ISV would lead to smaller and smaller increases. Additional products would 
minimize the effect of the z/OS increase. 

The analysis used prices in US Dollars. 

I did not assume the growth or pricing for CICS, DB2, or WebSphere MQ series. 
On the linked page above, in the lower portion I describe the actual increases 
of the version changes. I won't full repost here, but between DB2 V7 and DB2 
V10 the increases ranged from 9.5% to 12%. CICS TS V2 through V4 had a range of 
11.7% to 14.5%. Again over a range of MSU levels. 

I did not use zNALC pricing, this was VWLC and AWLC. I could run some zNALC 
scenarios. However, the effect of the z/OS increases would be even smaller with 
the very large zNALC discounts. 

In the US z/OS did have a price increase effective 1 April 2009 (see 
http://ibm.co/Dec2008-308-887-increases). I know as I typed those prices into 
the LCS pricing tables. For example the VWLC tier of 176 to 315 MSUs was 
announced on 10/3/2000 as $110, on 4/1/2009 this increased to $113 and now on 
4/1/2012 it will be $120. The lower tiers changed on 10/3/2003 with the 
mainframe charter; some tiers were raised and some lowered to as the initial 
z/OS announcement had a minimum starting base of 45 MSUs.

Thanks for all the comments and feedback, these are all good points in the 
discussion! 

Al Sherkow, I/S Management Strategies, Ltd. 
Consulting Expertise on Mainframe Software Pricing, 
WLC, LPARs and LCS Software 
Seminars on IBM SW Pricing 
Voice: +1 414 332-3062   
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Actually it is more typically 2% (Was Re: IBM announces 6% price increase for z/OS)

[Al Sherkow]

I realized I should have changed the subject. Now that I've done some analysis, 
when viewed with the full stack the increase is more typically 2% 

Al Sherkow, I/S Management Strategies, Ltd. 
Consulting Expertise on Mainframe Software Pricing, 
WLC, LPARs and LCS Software 
Seminars on IBM SW Pricing 
Voice: +1 414 332-3062   
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: IBM announces 6% price increase for z/OS

[Al Sherkow]

I've done further analysis, now that the LCS_PRICING Interactive Report is 
updated with the new prices. When viewed not tier by tier, but with a stack of 
other software products (for which the prices did not change) the increase on 
the MLC invoice due to the z/OS increase is closer to 2%. As more and more 
products are added to the software stack, the percent increase will be smaller, 
as the other products are added to both the before and after totals equally. 
One customer has shared their analysis with me and for their products and 
features the monthly impact is approximately 1.8%.

My analysis is available at:

http://www.sherkow.com/updates/zos_2012_price_increase.html

I hope you find this interesting and I welcome any comments or feedback you may 
have.

I can find no other public analysis that shows this effective increase is near 
2%. 

The bottom line is that everyone needs to analyze their own site. Similar to 
determining which software pricing metrics your site should be using there 
still are not clear rule of thumbs that work for every site.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Mainframe Software Pricing,
WLC, LPARs and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: zEnterprise Use Cases

[Al Sherkow]

There was a conference call and presentation to Analysts on Thursday. This info 
is from that call. 


-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on WLC
Seminars on IBM SW Pricing 
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: IBM announces 6% price increase for z/OS

[Al Sherkow]

I don't quite understand how the US Announcement (312-001) that clearly states:


Today, IBM® announces a price increase for the z/OS® operating system. 
Effective with the billing period that starts on or after April 1, 2012 the 
monthly license charges (MLC) will increase for the following program and its 
features:

Program number  Program name
5694-A01z/OS® V1


is already causing increased charges for some sites as reported on IBM-MAIN. (I 
don't mean to doubt those people, I would just like to understand). 

For full capacity sites the April 1st invoice should be the installed capacity 
(MSUs) on April 1, 2012.  For sub-capacity sites the February Data (Feb 2 to 
March 1) arrives at IBM between March 2 and March 9th. This updates the 
inventory for each sub-capacity product so that the April 1st invoice for 
sub-capacity products is based on the SCRT MSUs that have updated the 
inventory. 

Of course someone might have terms and conditions signed with IBM that allow 
any announced pricing increases to immediately take effect. Doubt that too. 
Perhaps some business partners are trying to get the increase early. Just seems 
strange to me.

Using the announcement letter as documentation of the effective date I think if 
I were advising those sites I would tell them to push back and not to pay the 
increase until the April 1st invoice. 


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Mainframe Software Pricing,
WLC, LPARs and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: IBM announces 6% price increase for z/OS

[Al Sherkow]

Here is some of what I wrote to clients and licensees on January 8th: 

--
IBM announced price increases for many features of z/OS including the Base 
feature on January 2nd, 2012. In the US the announcement letter number is 
312-001. (In Europe, Middle East, and Africa the announcement was December 30, 
2011 with letter number ZA11-1053.)

The February 2 through March 1st usage will be sent to IBM by March 9th, and 
these price increases will be included in the April 1st IBM MLC invoices. The 
increases are around 5% for the tiers that have been changed for AWLC, VWLC and 
ZELC metrics. If you are using Measured Usage and PSLC the z/OS increases are 
around 6.7%. For some z/OS features only the initial 3 MSUs in a PricingPlex 
were increased.
--

I was planning to post to IBM-MAIN once I had my analysis complete with some 
examples, but I haven't finished that work yet. It should not be retroactive, 
so I'm surprised that John has already had an increase. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Mainframe Software Pricing,
WLC, LPARs and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Re: Help: Delete LPAR definition from z9BC

[Al Sherkow]

John --

Another option that IBM could have pointed out  Your situation is very 
common, and IBM did modify SCRT to help sites with this issue. SCRT's PARMS DD 
statement provides a technique to hard code this information one time, so you 
don't need to do it month after month. In the PARMS DD input enter:

LPAR_Comment CPC=-s,LPAR=,Comment="ccc...ccc",
"ccc...ccc", ..."
ccc...ccc"

and the SCRT User Guide provides this example:

LPAR_Comment CPC=2096-12345,LPAR=LPARVM1,
Comment="This LPAR is only used for Linux"

which you can easily change for your old ICF LPAR. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on WLC, LPARs and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Frequency of Sysplex Verification Audits?

[Al Sherkow]

I'm curious about the actual frequency that sites are being asked to run IBM's 
Sysplex Verification Package. (The IBM Sysplex Verification Package can be 
found at http://ibm.com/systems/z/swprice/sysplex/verify.html .)

The Ts&Cs of the Attachment for IBM System z AWLC (this is in other documents 
as well) include this text:

-
"require Customer to provide IBMwith a completed IBMSysplex Verification 
Package upon the occurrence of any of the following: 1) Customer implements an 
actively coupled Parallel Sysplex, 2) a permanent change is made to any 
Eligible Machine, e.g., model upgrades or model downgrade in the Parallel 
Sysplex, 3) IBMrequests a package for any year during which Customer has 
received aggregated AWLC pricing, and 4) at least once in every 12 month 
period."
-

My impression based on my clients and licensees is that IBM is requesting more 
SVP audits than a few years ago. Have others been asked to do the SVP more than 
once per year? Especially larger customers with multiple machines the number of 
"permanent changes" in capacity could easily occur multiple times per year.

Thanks everyone!

Al

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062  
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Licensing of Tivoli Products.

[Al Sherkow]

The original post referred to "Tivoli Omegamon" which is an IPLA product. These 
products have a "one time charge" for each unit of capacity (i.e. for 
additional capacity you pay for additional OTC). Separately "subscription and 
support" is available for exactly that: support and upgrades to new versions. 
With S&S you do not have access to support and you are not entitled to new 
versions. 

If you want a new version and have not been paying the S&S I believe you need 
to "catch-up" on the S&S you didn't pay or you can pay for all the MSUs/Value 
Units again as if you had not acquired the previous version. You might be able 
to negotiate something with IBM. As far as I know all of my clients do pay the 
S&S, so I have no personal experience with a site not paying for that. 

As John wrote for z/OS, CICS, IMS, DB2, COBOL are paid for with a "monthly 
license charge" (MLC). You pay for these products every month. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: SMF timestamps

[Al Sherkow]

I'm using SMF90 subtypes 1 & 2 to catch time and date changes from SET 
commands. This has worked well so far to provide the logic necessary to 
identify that the clock has been set back and there will be SMF30s and almost 
every other record type with an hour duplicated due to Daylight Saving Time in 
the fall season. At least this lets me know the event has occurred. 

One issue I have run into with my product is that this record only has the 
4-char SMF System, not the 8-character SYSNAME or the SYSPLEX. (I have clients 
that have multiple 'SYSA's in their environment.)

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, and LCS Software
Seminars on IBM SW Pricing, and LPARs
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: IBM zIIP engine question?

[Al Sherkow]

Bob 

You can analyze the potential value of the zIIP engine by turning on PROJECTCPU 
in PARMLIB(IEAOPTxx). This switch has z/OS generate the "eligible" fields into 
the SMF data. If you're a member of SHARE or CMG you can find a paper I've 
presented a number of times: "zIIP and zAAP Eligible Time Analysis w/Workload 
License Charge Impact". 

-- 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Soft capping (Was: LPAR - MSU capping)

[Al Sherkow]

Kees- 

The LPAR's can belong to different sysplexes. Think about the data in the SMF70 
record that each LPAR is able to record about the other LPARs on the machine. 
That is the information that WLM has access to for managing the group. 

Each LPAR's WLM decides if that particular LPAR should be capped when the group 
MSU limit is exceeded or not. If the LPAR is over it's percentage of the 
group's capacity based on LPAR weight that LPAR will apply the cap. Not all 
LPARs in the group will necessarily cap together. 

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Soft Capping an LPAR

[Al Sherkow]

RMF MON III will report an estimated time until soft-capping will begin. This 
can also be monitored as an RMF exception condition so that you can set a 
threshold and take some action. 

If RMF III has this capability others monitors may also have it. I would 
contact the monitor vendor and ask them directly. Installing a new monitor (and 
paying for a new monitor) could totally wipe out the value achieved by soft 
capping. 

>From your description it does appear that the Defined Capacity may be too 
>small for the workload in this LPAR. It is a tradeoff between providing the 
>required service levels and the financial savings that the throttling 
>provides. When the LPAR is capped by either the soft cap or LPAR Group 
>Capacity Limits the impact within the LPAR is controlled by how WLM has been 
>implemented in the LPAR. From a WLM management point of view it is the same as 
>being short on CPU resource (the CPU being the only resource that is actually 
>limited by the capping). The work with the highest IMPORTANCE levels is 
>impacted first. (In this context high WLM importance indicated low CPU 
>priority). 

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Capacity

[Al Sherkow]

I can't understand from the original post, what the problem is that you are 
actually trying to solve?

Why do you want to have a Hard Capped LPAR in an LPAR Capacity Group?

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Google Groups Stopped "reading" IBM-MAIN

[Al Sherkow]

I usually monitor IBM-MAIN with GoogleGroups 
<https://groups.google.com/group/bit.listserv.ibm-main/topics> but that stopped 
showing updates on Friday, August 1st. 

My daily digest showed me I was missing the traffic, just thought I would post. 
(Though anyone else using GoogleGroups will also only see this in the digest. 

Guess I need to use another "reader". 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: WLC and 4HRA across CEC's

[Al Sherkow]

As I wrote to Peter on May 6: 

-
The Ts&Cs of WLC agreement signed by each customer include "the highest 
number of MSUs ... running on the same eligible machine during a reporting 
period."  

To my reading "same eligible machine" limits the simultaneous 4HRA to the 
"same eligible machine" not to all your machines. 
-

It's clear to me. 

John: You have not been overpaying due to this issue. (there may be other 
issues, but not this one)

Hal: The 4HRA is computed by WLM, and stored into the RMF70 data by RMF. 
SCRT works on the hourly level, not 30 minutes. 

I need to disagree with "while some (DB2, for example) charge for total MSU's 
across all LPAR's on that CEC (no matter if DB2 is installed on a given LAPR or 
not)". You are not charged for DB2 in an LPAR if it is not in the SCRT reports. 
SCRT does not know what you have licensed. It only reports based on the 
SMF89 data (on z/OS) and the NO89 control cards. 

I certainly agree with Hal on this point: It is worthwhile to get independent 
verification! 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: zIIPs and zAAPs

[Al Sherkow]

Bob --

I think the main point is acquiring the zIIPs with the machine avoids that 
potential future zAAP to zIIP conversion charge in the future. 

As Norman wrote with "zAAP on zIIP" all the work on that runs on the zIIP 
appears as zIIP work. The Java work is not "reported" in the SMF as zAAP any 
more. I'm not sure about the various zAAP "eligible" time fields (maybe someone 
else can clear this detail up). 

I think the odds are that overtime there will be more zIIP work in z/OS as IBM 
may create additional "eligible" workloads for the zIIP. Other vendors also may 
move some work to zIIP. Little additional work will go to the zAAPs. (though 
IBM has made some other work "eligible" for zAAP in addition to Java like XML 
processing). 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Dummy LPAR to store excess MIPS

[Al Sherkow]

Hi Timothy 

I haven't heard of that either. I was thinking of vendors that license for less 
than 
full capacity for a specific number of MSUS or MIPS. 


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Dummy LPAR to store excess MIPS

[Al Sherkow]

LPAR Group Capacity Limits came with z/OS 1.8, so that should be available to 
most sites. Requires z hardware. What some customers have not understood is 
you are allowed to have multiple LPAR Groups on a single machine and they are 
able to work across sysplex boundaries. 

So some sites use one group to limit a whole machine to 80% of installed 
capacity 
regardless of the number of LPARs. 

I have other sites that have one group for Production LPARs, a second for 
Development LPARs and a third group just for QA LPARs. 


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Dummy LPAR to store excess MIPS

[Al Sherkow]

Timothy --

Yes, Defined Capacity and Group Capacity Limits can do this, but not *all* 
vendors. For many vendors, but not for all. Hence some sites are using these 
other techniques. 

Al


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Integrated Workload Pricing

[Al Sherkow]

Yes, I will be talking about this at Share as Bob pointed out:

"Changes in Workload License Charges with the zEnterprise"
Tuesday, March 1, 2011: 4:30 PM-5:30 PM 
Room 204BC (Anaheim Convention Center) 

This certainly qualifies as a change. During the week of Share will be the 
first 
time that customers submit SCRT reports that would reflect this pricing option. 
It could be important to have SMF APAR OA31615 installed by before February 
2nd if you have LPARs with CICS or IMS and any of the 'IWP Defining Program' 
that are listed in the announcement letter. 

I do have some test data and I'll be studying the potential impact on in the 
next week. I'll hope to have real examples to use at Share. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Integrated Workload Pricing

[Al Sherkow]

Tuesday, 18 January 2011, IBM announced Integrated Workload Pricing (IWP) 
that may reduce the billable MSUs for CICS and IMS if they are executing in 
LPARs that also are running certain other WebSphere products. The 
announcement letter is number 211-011 with a full title of 'IBM Integrated 
Workload Pricing for zEnterprise clients can significantly improve 
price/performance of collocated workloads on z/OS'. Further prerequisites are 
z196 servers, and z/OS 1.10. The lists of CICS, IMS and WebSphere products 
are in the announcement letter at these hyperlinks:

http://tinyurl.com/5w8g6xt 

http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?
subtype=ca&infotype=an&appname=iSource&supplier=897&letternum=ENUS211-
011 

IBM refers to the WebSphere programs as 'Integrated Workload Pricing Defining 
Programs.' At this point all the 'Defining Programs' are of IBM's 'WebSphere' 
brand.

You may have heard me discuss or present that IBM's Variable Workload 
License Charges (WLC) can be thought of as 'Container License Charges'. It is 
the size of LPARs as containers of software products that determines the 
charges for a product. (WLC has nothing to do with the traditional concept of 
mainframe 'workloads' or the 'workloads' of the z/OS Workload Manager).  This 
announcement makes the 'container' smaller for CICS and IMS by determining 
the 4 hour rolling average of  WebSphere products based on their own CPU 
time. This is the CPU time of these 'Defining Programs', as recorded in the 
SMF89 data. The 'Defining Programs' 4HRA is subtracted from the traditional 
4HRA of VWLC for the LPARs. The smaller container of IWP is only for the CICS 
and IMS products in the announcement letter. Through IWP the billable MSUs 
reported by SCRT for the CICS and IMS products should be reduced, leading to 
reduced software charges. Any IPLA Reference-based products related to CICS 
or IMS as parents will have a reduction in the number of Required Value Units.

Integrated Workload Pricing does not change the charges or value unit 
requirements of the WebSphere products. The 'value' is in reduced software 
charges for CICS and IMS. 

SMF APAR OA31615 is required for some of the WebSphere 'Defining Programs'. 
It provides enhancements to the SMF89 data. An additional area of 
enhancement in this APAR is inclusion zIIP and zAAP CPU time which are added 
together and referred to as the 'Product Offload Engine Time' in the SMF89 
data.  

At this point I haven't yet analyzed real data to provide an example. If you 
have SMF APAR OA31615 installed I'm looking for some test data! 

Best regards, Al 

Al Sherkow, I/S Management Strategies, Ltd. 
Consulting Expertise on Capacity Planning, Performance Tuning, 
WLC, LPARs, IRD and LCS Software 
Seminars on IBM SW Pricing, LPARs, and IRD 
Voice: +1 414 332-3062 
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/os x z/os.e differences

[Al Sherkow]

I don't think you'll be able to IPL zOS.e on a z10BC as that message indicates. 
Also I don't think IFAPRDxx parameters will allow you get around that. Apply 
for a 
zNALC qualification working with your local IBM reps or business partner. 
Carlos it 
looks like you work for IBM so that should help in this. 

If the workloads qualified for z/OS.e in the first place getting zNALC 
qualification 
should not be difficult. 

Al Sherkow
Self-Proclaimed Independent IBM Software Pricing Expert

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: CPU capping is not working for one Lpar only on CEC?

[Al Sherkow]

That 3% is not unexpected. WLM works with PR/SM to implement the cap. WLM 
does the math and PR/SM does the actually "capping". If you read the PR/SM 
planning guides (as Peter referenced) you'll see that PR/SM manages LPARs to +-
3 percent. 

I believe that WLM does the calculation to account for the possible MINUS 3%. 
If 
you were trying to cap at 100 MSUs and you only got 97 MSUs while paying for 
100 you would be upset. If you pay for 100MSUs and get 3 Bonus MSUs you are a 
satisfied customer. (I chose 100 MSUs to make the math easy)

Al Sherkow, I/S Management Strategies, Ltd. 
Consulting Expertise on Capacity Planning, Performance Tuning, 
WLC, LPARs, IRD and LCS Software 
Seminars on IBM SW Pricing, LPARs, and IRD 
Voice: +1 414 332-3062 
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: CPU capping is not working for one Lpar only on CEC?

[Al Sherkow]

It is possible for the 4HRA of an LPAR to go above the defined capacity limit. 
This 
is because the limit goes on typically after four hours of your workload 
getting 
bigger and bigger. Eventually the 4HRA exceeds the defined capacity and the cap 
goes on. As the four hour rolling average goes forward, the values from 4 hours 
ago, then 3h45m ago, then 3h30m ago are removed from the calculation of that 
average, and values at the capping level are going into the calculation, so the 
4HRA may keep rising for a few intervals. It may even stabilize above the set 
defined capacity. Back in 2004 I coined the term "bonus MSUs" for these MSUs 
that are above the defined capacity value. When SCRT processes your SMF data, 
if the 4HRA in the data is above the defined capacity value SCRT ignores those 
"bonus MSUs". The Ts&Cs of sub-capacity WLC indicate you will not be charged 
for more MSUs than the Defined Capacity Value. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Soft Cap

[Al Sherkow]

I've presented on this fact at Share conferences, z/OS Expo and CMG. The 4
hour rolling average (4HRA) often (but not always) stabilizes at a number
higher than an LPAR's Defined Capacity or an LPAR Group's Capacity Limit.
This has to do with the algorithms WLM uses to control the capping. 

When you set or lower a capping target (either Defined Capacity or LPAR
Group Capacity limit) it of course cannot impact the history that has
occurred in the preceding four hours. The 4HRA continues to be computed
based on that history and may well be above the capping value. However, when
SCRT processes the SMF data the 4HRA, and the defined capacity and or the
group capacity are all in the SMF 70 data and SCRT will report the lower of
the values. So you 4HRA can be higher than the cap and you are not charged
for that. 

When I first presented this in the Spring of 2004 I called this effect
"Bonus MSUs" and that was a term IBM certainly liked!

If you are capped is at 16 MSUs then SCRT should not be reporting 17 MSUs.
This indicates that at least one hour in the month the 4HRA of your LPAR
group was at 17 without the cap being enforced. This is one of the features
of the LCS software: auditing of SCRT.  

Al Sherkow, I/S Management Strategies, Ltd.
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

z196, AWLC updated information

[Al Sherkow]

It's public today. The question was asked this morning at Share. The z196 
tech dividend is provided through the new AWLC pricing metric & NOT 
separation of HW MSUs & SW MSUs. I've updated my analysis of the z196, AWLC 
and their impact on software pricing. Please see 
http://www.sherkow.com/updates/z196awlc.html.


--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web:www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

New zEnterprise 196 Hardware, AWLC and IBM Software Pricing

[Al Sherkow]

Cross Posted (MXG-L, ISVCOSTS, IBM-MAIN)

Today IBM announced the zEnterprise 196, a new mainframe system 
available in September that will also offer a BladeCenter extension 
available in November. The US announcement letter is 110-170. The z196 
will also have a new pricing metric "Advanced Workload License Charges" 
described in US announcement letter 210-238. AWLC offers "Reduced price 
points, versus VWLC, for all MSU levels with more than three MSUs". This 
will help lower IBM MLC software charges for sites using "VWLC" (but no 
help if you're using "EWLC").


Further, AWLC will take advantage of tiers that only z/OS uses today. 
DB2, CICS, WebSphere MQ , IMS and other MLC WLC products do not lower 
the charge per MSU above 575 MSUs. For Example DB2 V10's charge at 575 
or more MSUs is 59USD. With this announcement I'm expecting the 59USD to 
be decreased ("Reduced price points, versus VWLC"), and there should be 
further decreases for levels 5-7:

 Level 5 AWLC 876 - 1315 MSUs
 Level 6 AWLC1316 - 1975 MSUs
 Level 7 AWLC1976 - more MSUs
   
The prices have not yet been published, but each tier should have a 
lower price than 59USD per MSU. This will lower IBM MLC software charges 
for sites with more than 875 MSUs in their PricingPlexes. In my opinion 
this is long overdue. A 2094-715 or a 2097-711 or larger machines will 
benefit from decreases at these higher tiers.


Further updates on z196, AWLC and their impact on software pricing will 
be on http://www.sherkow.com/updates/z196awlc.html.


Al

--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: How low can you go?

[Al Sherkow]

Ken you're correct; there is a minimum charge of 3 MSUs for an IBM product
per "PricingPlex". That's important because it is not per machine. 


Al Sherkow
I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: How to be notified when softcap is set?

[Al Sherkow]

David --

There are no z/OS (WLM, RMF) messages when the cap is reached.

You can monitor this with RMF III from the CPC screen. I don't know about
the other monitors, but I expect they would have a similar capability.  


Al Sherkow
I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Softcapping and service goals

[Al Sherkow]

David --

All the measurements and goals work as if softcapping was not active. When
the cap is on, as you wrote, you are in control based on your service
definitions. When softcapping is in effect, you are "short of CPU", and the
lower importance work will slow down. That starts with "discretionary",
which is just that, work to run when the resources are available. While
softcapping the resources are not available. 

While WLM works the same, the consumption of CPU by other service classes
may decrease also; this lower CPU time and lower rate of consumption may
have other impacts on your definitions. 

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: LPARs: More or Less?

[Al Sherkow]

With z/OS and sub-capacity pricing the charges for many important and key
products (and the more expensive) are tied to the "size" of the LPAR. A
small amount of IMS in a large LPAR is charged a larger price than a small
amount of IMS in a small LPAR. 

Some of IBM's current policies cause sites to jump through hoops to save
money. This is one area, and another is parallel sysplex aggregation. When
parallel sysplex was introduced the aggregation was a financial incentive to
move sites to parallel sysplex. Most sites are there by now :)

But sites are being forced to follow the rules initially set long ago. That
is not a productive use of valuable system programmer time. Some sites spend
a lot of effort managing this. They could be installing a new release of DB2
or IMS. They could be working with linux on z, or websphere or any number of
other strategic initiatives for their organization. Instead they are jumping
through hoops to follow rules that may lower their software charges, but do
not enhance what mainframe if providing to their business. 

In general, and YMMV, more LPARs leads to lower software charges. The
various LPARs do not all have their peak 4HRA during the same hour. The
simultaneous 4HRA across all LPARs is lowered. In general I don't advise
breaking apart LPARs just for SW charge savings, but in some cases the
savings are significant. The IMS example in my first paragraph is a good
example of that. I do recommend that sites stop combining LPARs for easier
management. Rather leave them separate for lower SW charges, again YMMV.
This has been a fairly standard practice after a merge to combine for
example two separate production LPARs into one large Production LPAR. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: cpu capping question

[Al Sherkow]

Tommy --

Yes, LPAR 1 has access to the capacity that LPAR 2 is not using. The cap on
LPAR2 is a limit only on LPAR2's use of your capacity. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: ibm.com/systems/z/index.html is broken

[Al Sherkow]

seems to be working now. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

ibm.com/systems/z/index.html is broken

[Al Sherkow]

Just a heads up that some portions of the IBM.COM website are down! Techdocs
and the portions of IBMLINK that I use are working, but much of the
reference material is not available. 



Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Shop zSeries Ordering

[Al Sherkow]

I had numerous problems with IBM.COM yesterday. I was searching for
information on the products added to SCRT V18 and could not find them. For
example 5655-V60 is WebSphere Message Broker for z/OS V71 (actually only V7,
not 71) and the only hit on IBM.COM is the SCRT What's New Page. Fortunately
a google search was able to find the announcement letter so I could gather
the other information I need. I've always been able to find announcement
letters via the IBM search box. The search doesn't work today either. 

Al Sherkow, I/S Management Strategies, Ltd.
Call Save Money on IBM System z Software Charges
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Just installed a new IBM z10-BC

[Al Sherkow]

Two factors here. First, the zIIP was added to the environment and a portion
of the DRDA work moved from the general purpose engines to the zIIP. Second
the technology changed from 2096 to 2098. Between those two machines there
is what IBM refers to as a "technology dividend", roughly 9-10%. A software
MSU on the new machine does more "work" than a software MSU on the old
machine. Even without a zIIP, if there is no latent demand in the workload
moving from a 2096 to a 2098 would lower your old 110 MSUs to 9 or 10% less
MSUs. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Seperate LPARs for Prod and Test or a single LPAR for both

[Al Sherkow]

>From a software pricing point of view separate LPARs are often cheaper. Some
products are only needed in development (compilers, some development tools,
some testing tools), most (if not all) products you need to license for
production you'll also want to use on your test LPARs. (otherwise how would
you test new releases and maintenance). For products that are licensed
without sub-capacity, at the machine level, this does not matter (though you
might have benefit isolating Testing to one of multiple machines). 

It's possible, though in my experience with many shops, not too likely, that
production LPARs and testing LPARs do not have their 'peak's at the same
time. So that in general separate LPARs are better than combined LPARs. 

Of course, your mileage may vary. 

Al 
-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z Training: IBM System z Expo - Oct. 5-9, 2009 - Orlando

[Al Sherkow]

There are definitely user speakers at zExpo this year. I know that one of my
clients is presenting two sessions. One on zIIPs and one on LPAR Group
Capacity Limits. Both of these are User Experience Sessions. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/OS SCRT NO89 Product Information

[Al Sherkow]

I believe the fine print Ts&Cs indicate that if you specify that a product
runs on one LPAR or any LPARs of a machine for which it is not licensed that
IBM may view that as an order for the products. 

Similarly, for products that do generate SMF89 data, if SCRT detects a
product on a machine where it is not licensed, that is an order for the
product. 

You can correct this after the fact, but it's better to read the reports and
be sure they are what you expect. (LCS highlights the discovery of products
running where they are not expected to be running based on your licenses
and/or history of product usage). 

This is the basis of Pat's original question. You are supposed to setup the
NO89 parameters to reflect in which LPARs you actually use the NO89
products. This is why LCS detects this, to help sites properly report their
usage to IBM. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/OS SCRT NO89 Product Information

[Al Sherkow Digest]

Setting the NO89s to *ALL will just indicate to SCRT that you are running
those products in every LPAR on every machine. Probably true for products
like NetView, IBM's System Automation and their Scheduler. Lots of products
once you commit to using them you need to run them everywhere. But for
COBOL, PL/I and development/testing products that may not be true. 

I don't think you'll "learn" anything by trying *ALL. 


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/OS SCRT NO89 Product Information

[Al Sherkow]

I can't resist jumping in here. Especially since the Patrick included "Al,
help!?" in his original post.

I think it's true as Mark wrote that "Tivoli License Compliance Manager" can
help with this task. But in the US that's $2000 per MSU for the OTC alone
with the VUE007 conversion table. TLCM will do a lot of things LCS does not
attempt to do but LCS is what you need to audit, manage and optimize for SCRT.

LPAR Capacity and Software Usage Analysis (LCS) Software is only
$15,000/site. With TLCM $15,000 would only license 7.5 value units, only 15
MSUs of capacity. (Patrick also wrote "quite a few physical machines and
about triple the amount of LPARs" so 15 MSUs probably isn't enough). The ROI
on LCS is often achieved with next month's SCRT report. How many products
can claim a one month ROI!

LCS will identify the NO89s that typically not used in every LPAR like
COBOL, PL/I. LCS will even generate the NO89 control cards for you. Read
more at my website.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Subcapacity Planning Tool

[Al Sherkow]

It's meant to be a quick tool to help you plan. I don't think it was
intended to help you manage your machine. Did it start reporting when it had
4 hours of data from all LPARs again? 

Al 

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Offload work to ziip

[Al Sherkow]

I’ve worked with many ISVs, but typically NDAs are signed and I am not
allowed to even disclose a relationship exists. I have been consulting with
Neon Enterprise Systems in relation to zPrime. I am an independent expert
and the information in this IBM-Main posting is no longer under my NDA.

Neon Enterprise Systems is not using the API in zPrime. They have invented a
new way of enabling work for the zXXP engines and that is a 'trade secret'
of Neon Enterprise Software. 
 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com <http://www.sherkow.com>

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Adding CPU engines without IPL

[Al Sherkow]

Hi Jim --

I think the engines just be in the LPAR configuration as 'reserved' at the
previous IPL for MVS to recognize them. PR/SM can use additional engines
lowering the logical to physical ratio, but MVS only builds enough control
blocks on for the number of initial engines plus the number of reserved
engines. 

This may have been improved with recent releases of z/OS but this was
certainly an issue in the past.

Regards,

Al
-------
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
zPricing, LPARs, IRD and LCS Software

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Another Group Capacity question

[Al Sherkow]

Set a defined capacity for the development LPAR to keep this possibility
from happening. 

Hal's points are correct also. It just does not work the way you've written.
The key is that it is a "rolling" 4 hour average. 

Al Sherkow
LCS software and Consulting for z Pricing

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z9: Group Capacity and weights.

[Al Sherkow]

You've got it correct

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Soft capping advice sought

[Al Sherkow]

You don't need to set Defined Capacities for the other LPARs (though you
may). If you want to limit the 4HRA of an LPAR to limit what that particular
LPAR can contribute the 4HRA of it's products you can. If you don't card
about the individual LPARs don't set a defined capacity. The Group Capacity
limit will limit the whole group. 

For example let's say IMS/DB is in only in LPAR C. If you want to limit the
IMS/DB charges to xx MSUs then set the Defined Capacity of LPAR C to xx.
Otherwise IMS/DB will be billed on LPAR C max 4HRA within the Group. 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Soft capping advice sought

[Al Sherkow]

Put them all in one group, set the group cap to 100

Set LPAR A's defined capacity to 100, and give A the most weight in these 5
LPARs.

Leave LPAR B with it's defined capacity of 10.

Leave C,D and E with less weight than A & B. 

This will let A take all 100 MSUs if it wants that much. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Performance problems

[Al Sherkow]

As you've been told set Defined Capacities for your LPAR(s), and/or if there
are multiple LPARs an LPAR Group Capacity Limit.

Of course you must be on an IBM sub-capacity pricing metric. 

If your prime shift is really at 20%, then set the cap very low. This will
slow down the batch and help control the cost. If you're on sub-cap then the
3 hours of batch at 100% is driving your bill. 3 hours at 100% and 1 hour at
5% (a guess for an idle system) has a 4 hour rolling average of 305/400=76%.
If you limit it you can lower your costs. 

You might consider an independent consultant to help with the pricing
analysis... 


Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Consolidation

[Al Sherkow]

I won't comment on the technical question you asked. 

However, there is a non-technical reason for a "primary" parallel sysplex.

There would likely be some significant savings in your monthly IBM software
bill if your three sysplexes were merged so that you had one "primary"
sysplex. This does not mean all images must be in a common sysplex, but
rather the majority of the prime shift CPU time across the 3 machines must
belong to one primary sysplex. (You can of course have multiple images on a
machine in this primary sysplex.) 

The rules are non-trivial, so besides referring to the IBM sysplex
aggregation pages <http://ibm.com/zseries/swprice/sysplex/> you might want
to review the 76 page Red-Piece "z/OS Systems Programmers Guide to: Sysplex
Aggregation" <http://www.redbooks.ibm.com/abstracts/redp3967.html>,

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Limit and Weights

[Al Sherkow]

The capacity of the group is distributed to the LPARs in the group based on
the normalized weights of the LPARs in that group. 

When the group limit is exceeded L1 would be 40 MSUs, (80% or 50) and L2
would get 10 MSUs. If L1 does not require 40, then the excess can be used by
L2 (just like without group limits). If L1 is not running at all then L2 can
have all 50 MSUs of the group. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Capacity Limit

[Al Sherkow]

Greg is correct, that SCRT reports the maximum simultaneous 4HRA for each
product. That the LPARs have individual peaks at separate times during the
month is not an issue. 

I also agree with Hal that soft capping and group capping are more business
related than technical related. (I have a few counter examples in my seminar).

So I disagree with Errol's plan. 

Defined Capacity limits how many MSUs a particular LPAR can contribute to
the maximum simultaneous 4HRA for each product on the machine. 

The Group Capacity Limit is a throttle on how many MSUs a particular LPAR
can contribute to the maximum simultaneous 4HRA for each product on the
machine. An example of this is to limit *all* the testing LPARs to a
specific number of billable MSUs. A 2nd example is to limit departments or
clients LPARs to a specific number of MSUs. I have some customers with
multiple groups (Test, QA, Production) and other customers with one group to
limit the billable MSUs of the whole machine.

KEY POINT: Neither has any impact of which LPAR can access the excess
capacity of which other LPAR. The concept that LPAR T will use the excess
capacity of LPAR P rather than the excess capacity of any other LPAR on the
machine is not correct. 

Both defined capacity and group capacity limit an LPAR in the future based
on the last 4 hours of history. In other words the use of excess capacity in
the next interval (or the next 10 seconds or 30 seconds or 2 minutes) is
based on what has already occurred. If you want to limit LPAR T from ever
getting too much capacity then hard capping is probably the answer.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Capacity Limit

[Al Sherkow]

I did not see any mention in this thread of more than two LPARs, or "many
LPARs and many different customers on the Processor complex", or that you
were using softcapping on any LPAR "LPAR P typically gets soft-capped about
midday".

So I answered two and only two LPARs with no capping of any kind on a single
machine. 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Capacity Limit

[Al Sherkow]

Answer cross-posted to LPAR-PRICING-L and IBM-MAIN

As I understand your question you want LPAR P to use the excess capacity of
LPAR T but not vice versa. 

Does this have anything to do with Software Pricing??

LPAR T can be limited to not using P's excess with traditional LPAR hardcapping.

If you want to limit the MSUs that LPAR T can contribute to the simultaneous
4HRA of that machine then LPAR T can be controlled with Defined Capacity.
LPAR T can access the excess from LPAR P, but will not be 'charged' for that
access. 

The only reason to use a group is if you want to limit what the LPARs of the
group can contribute to the simultaneous 4HRA of that machine.

IRD is not involved or required. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z10 and HiperDispatch question.

[Al Sherkow]

That is how HyperDispatch should work. Engines will be pseudo dedicated to
an LPAR. 3 LPARs on a W03 (or U03 of the orig question) with 3 logicals each
would tend to have one engine at 100% and the other two close to zero (if
all have equal demand) and weights of 33%.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z10 and HiperDispatch question.

[Al Sherkow]

on a z10BC U03 you probably won't see a 'measurable' difference. All the
LPARs are in a single book and all the physical engines of an LPAR will be
close together already. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Need to cap my software costs

[Al Sherkow]

You are correct that you do not have LPAR Group Capacity limits with z/OS 1.7.

You can cap those small LPARs but you must use a soft cap on the 'most loved
LPAR' also. Otherwise that will drive the box to 100% (if there is workload
that wants that capacity). 

The max 4HRA for invoicing is simultaneous across the LPARs within any hour.
The limit of this number is the sum of the defined capacities. You might
want to both hard cap and soft cap the small LPARs.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: WLM four-hour rolling average

[Al Sherkow]

The 4HRA does not change with changes in Defined Capacity. If you raise a
defined capacity value, while WLM is limited the LPAR due to implementing
the defined capacity, then the 4HRA will start raising also. But this is a
four hour average, so it is not a large jump, and certainly not a 'resetting'. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: IBM Debug Tool experiences

[Al Sherkow]

BTW: Debug V9 is IPLA only, where Debug V8 was MLC.

If you have the licensed the IPLA add-on product Debug Tool Utils & AF V8
(5655-S16) then the S&S on that allows you to add Debug V9 without
additional OTC. If you don't have that product, you'll need to pay the OTC. 

There is a good side, If you have both Debug V8 and Debug Tool Utils & AF V8
and you pay for the S&S when you move to Debug V9, there is no more MLC and
the OTC you've paid for Debug Tool Utils & AF V8 S&S provides you with Debug
V9 (which includes the Debug Tool Utils & AF). 

Debug Tool Utils & AF is no longer a separate product.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z10 power problem notification

[Al Sherkow]

Based on the original discussion on MXG-L I wrote a email to my clients and
licensees on IWM063I and it potential impact on your software charges. You
can read it at 

http://www.sherkow.com/updates/20081014cooling.html

There are circumstances when this is a normal message. Such as when changing
the speed of a machine that supports capacity settings. But you'll know when
these circumstances are happening in your shop. If you receive this mention
and you weren't expecting it then you have a problem! 

The message is issued again when the problem is fixed.

Regards,

Al Sherkow, 
I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/OS V1R10 & COBOL

[Al Sherkow]

There is a 28-page PDF on the upgrade options to get to Debug Tool V9. See


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/OS V1R10 & COBOL

[Al Sherkow]

and ... ... if you need the Debug Tool with COBOL V4 beginning with Debug
Tool V9 (5655-U27) it is now an IPLA product. Previously Debug Tool was MLC.
There is a separate add on product called Debug Tool Utilities and Advanced
Functions, and that has been IPLA. 

If you have Debug Tool Utilities and Advanced Functions you get the new
DEBUG TOOL as part of the S&S on the other product. If not, when you want V9
you need to pay the One Time Charge and start the S&S. 

The US Announcement Letter is 208-264 dated 23Sep2008.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: WebSphere 7.0

[Al Sherkow Digest]

It was announced 9Sep2008, US announcement letter: 

http://www.ibm.com/common/ssi/rep_ca/8/897/ENUS208-208/ENUS208-208.PDF

and 

http://www.ibm.com/software/webservers/appserv/was/

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Chargeback for zIIP & zAAP Usage

[Al Sherkow]

Ted --

I also would lean towards "C. Some sort of discount". 

I also think the rate for general CPU could be lowered due to zXXP engines
(I borrow Kathy Walsh's shorthand). Your *users* do not necessarily have
control over whether there work goes to zXXP or not. Some of this depends on
how many zXXP you have installed and available in LPARs relative to the
amount of work that could use the zXXP. 

If you lower the general CPU rate you are passing on some of the platform's
lower TCO with zXXP to *all* the users/customers. 

If you carry this far enough you lower the general CPU rate, and then you
could use your option "A. The same as for regular CPU."!

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Explaining the New IBM "Getting Started" WebSphere Pricing Announcement

[Al Sherkow]

IBM is doing a webcast on "Getting Started Sub-capacity Pricing for z/OS
IPLA Software" on Wednesday. Here is the link http://ibm.com/software/systemz/webcast/30apr/";>
http://ibm.com/software/systemz/webcast/30apr/

Al Sherkow
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: MULC availability

[Al Sherkow]

ULC is absolutely available with PSLC on the newest machines. 

IBM might "prefer" that you use Workload License Charges, but in fact there
are sites that are better off without WLC and that is often because they are
using ULC and Parallel Sysplex License Charges (PSLC). You cannot use ULC
with WLC, so those sites stayed with PSLC.

Contact your independent consultant for an independent analysis of which of
the alternatives is best for your site.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Explaining the New IBM "Getting Started" WebSphere Pricing Announcement

[Al Sherkow]

This is my opinion only. IBM is doing this to make these WebSphere products
more attractive for implementation on zSeries. Why not make these straight
VWLC? well lots of sites have bought value units and don't want to pay
monthly for these again. 

This is a guess, because the announcement is quite vague about how this will
work, but the only way SCRT can measure a product within an LPAR using the
SMF70 and SMF89 data is through the CPU consumption data that is already in
the SMF89 data. They are doing now for MQ Series with the SALC metric. So
that precedent already exists. In effect like SALC this is another type of
measured usage within sub-capacity pricing that does not allow measured usage. 

"Oh what a tangled web we weave" Sir Walter Scott, Marmion, Canto vi. Stanza
17. Scottish author & novelist (1771 - 1832)


Al Sherkow
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and *LCS Software*
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Explaining the New IBM "Getting Started" WebSphere Pricing Announcement

[Al Sherkow]

Timothy --

Thanks for the nice explanation. 

Besides using SCRT the site also most be sub-capacity for IPLA. This is a
separate agreement with IBM. If you take a snapshot today of your
environment to determine if you should use Sub-Capacity IPLA the answer will
likely be that you cannot save any money. If you are already properly
"entitled", that is you have the proper number of value units for your IPLA
products, then Sub-Capacity does not matter. 

And as your workload grows you still have the proper number of value units
because you acquired them for the "whole machine". When it is important is
before the *next* upgrade. If you go to sub-cap IPLA before the next upgrade
you may have enough entitled value units to not acquire more with the
upgrade. Do not expect your sales person to point this out to you! 

You may have excess value units, but you must still pay the S&S until your
use grows into them. But with those excess value units you can use the IPLA
products in additional LPARs or on additional machines. Your value units are
the for enterprise, not a specific machines. This is true of any IPLA
product not just the websphere family. 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: SALC pricing for MQ

[Al Sherkow]

You are both right, but you're missing point of IBM's SALC metric. There are
still quite a few sites that cannot save money with IBM's Sub-Capacity
pricing. You are currently paying ULC (Usage License Charge) or MULC
(Measured Usage License Charges) probably along with PSLC (Parallel Sysplex
License Charges). 

For many of the sites still using PSLC they are using a combination of PSLC
and MULC. This MULC applies to some key products such as DB2, CICS, IMS, and
importantly MQ Series. If you use a small amount of processing power you may
be able to save money with MULC. Many sites have MQ licensed this way
because the one MSU each of you is using is certainly much less than paying
for your installed capacity as otherwise is the basis of PSLC.

Now on to Sub-Capacity Pricing with Workload License Charges.

Once you start WLC you can no longer use MULC. So if you had DB2, CICS, IMS,
MQ on MULC with WLC they will be charged based on the 4 hour rolling average
of the LPARs where they run. Some sites only had MQ series on MULC, and they
still could not save money with WLC because of the increase from they small
usage of MQ in CPU time, to the 4HRA of the LPARs. And if you use MQ it is
often in most of your LPARs. 

As you wrote $10,120 is SALC charge for 1 MSU of MQ's CPU Time. 

With WLC pricing for MQ V5, 123 MSUs of 4 hour rolling average in the LPARs
running MQ the charge would be $10,108. 

So the point of SALC is to allow more sites to move to WLC and have a
smaller charge for MQ than the WLC charges would lead to. 

This is one of those points that IBM may not tell you about. Remember
which consultant told you about this when you want an independent analysis! 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Specialty Engine Prices (Was: Another RNL question)

[Al Sherkow]

It certainly was hard to find, but I finally did find a reference to
carrying forward your specialty engines when you get a newer technology
machine. Here is the full path (watch the wrap):

ftp://ftp.software.ibm.com/software/ae/events/doc/system_z/1
0_when_the_mainframe_is_the_cheapest_v3.9_for_europe_q1.pdf

or with a TinyURL: http://preview.tinyurl.com/695er3

The top chart on page 16 has "free upgrade to next generation".

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: MVS initiated FTP to Windows

[Al Sherkow]

I run an ftp server on windows and it works quite well. Someone suggested a
google search, the one I use Serv-U (http://www.serv-u.com/). And as Ed
wrote I use both DHCP and DDNS. 

Al

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: LPAR Rename

[Al Sherkow]

Besides the HCD manuals this is also described in the various hardware
redbooks like: "IBM System z9 Business Class Technical Introduction".

I think RMF/SMF70-1 will only have 'activated' LPARs regardless of being
IPL'd or type of LPAR (MVS, IFL, Coupling Facility). 

So '*' can be a place holder, but not activated. 

Al Sherkow

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Sub-CEC Reports

[Al Sherkow]

There has been a 'request' in to IBM for some time related to this
automation/oops/ no automation topic. 

The request is: allow sites to FTP or email the reports to IBM. IBM would
then stage your reports for your comments "as-if" you had used the web
interface. That is, they would get ready for you to review and comment on
your reports, and not assume that they are "ready" for final processing. 

This avoids the issue of using the LMDS web application to upload your
reports. It is fairly straight forward as this thread has shown to email or
FTP out of your monthly batch job. 


If your company does not review/update/verify by the 9th of the month when
the reports are due then IBM would process them. (This handles the "oops"
case where your automation sends the reports, but no one at your company
looks at it). 

I don't know the status of this request. If you don't send the report at all
IBM is entitled by the T&Cs of your sub-capacity agreement to bill you for
the installed MSUs of each machine. 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Sub-CEC Reports

[Al Sherkow]

Directly sending the SCRT reports to IBM without a review has some risk
associated with it. The SCRT report will result in an IBM calculating your
software invoice for hundreds of thousands or even 1M+ dollars per month (or
whatever your currency is). I recommend you examine the report for a number
of reasons: 

Does it have the numbers you are expecting?

Does it have the products you are expecting?

Did someone start a product incorrectly (on the wrong machine or in the
wrong LPAR or an incorrect version)? 

Does the report meet the 95% "Data Collected for z/OS" guideline?

Does the report include CBU testing MSUs that should not be included on your
invoices?

Does the report include an "unusual situation" that you should change by
specifying a smaller number in the "Customer MSUs"?

There are other reasons, but *each* of these has occurred among my clients
in the past 3 months.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: No more text messaging from mainframe?

[Al Sherkow]

Hal --

The problem I had with cingular/at&t was with the accounts for the phones.
My phones were set not to accept text messages (because now i need to pay
for them and I don't have a messaging plan). so I had them add the 'default'
which is pay per use, and now it works. 

Previously the messages were not delivered *and* they were not returned as
undeliverable (or user unknown, or anything useful). they just disappeared.

so check the plan for your phone. 

good luck

Al 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: IPL an LPAR with a very low weight?

[Al Sherkow]

Hi Dave --

I suspect you wouldn't be considering this unless you were having problems
with CPU resource contention in the two big LPARs (at 47 and 43). What kind
of machine is this? how many engines? 

If this small LPAR is down, and the box is nearly 100% busy, IPLing this
LPAR will be very slow, and this small LPAR may slow down the others in the
sysplex and MIM.

What is the real problem you are trying to solve? 

Al

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z800 to z9 migration

[Al Sherkow]

Hi Ted --

That may be true, but John has one CP and is looking at 2 CPs, so straight
lining at 24 CPs is far in his future. Also he wants free or low cost. 

Respectfully,

Al

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z800 to z9 migration

[Al Sherkow]

Look into zPCR from IBM. Just the tool you need. Start with this link
<http://www.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/PRS1381>. 

Best regards,

Al

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: WLC, WLM and all the rest: how optimize them ?

[Al Sherkow]

When you have signed a sub-capacity IPLA agreement, then your sub capacity
products are 'tied' to a different number than your installed capacity. As
someone pointed out earlier <http://ibm.com/zseries/swprice/zipla/> is the
best reference for customers. 

There are different types of IPLA products, that can be found on this page
<http://ibm.com/zseries/library/swpriceinfo/ipla.html>. In general, the DB2
related products will be charged based on the DB2 MSUs, CICS products based
on CICS MSUs, and the same for IMS. Some products record SMF89 data and they
are changed as other Variable sub-cap products (WebSphere, Fault Analyzer,
App Monitor, etc). Still others are charged based on z/OS MSUs rather than
the full box. 

If you assume your licenses are correct today, then this becomes really
important before the next upgrade. All the MSU numbers are computed to Value
Units. Let's consider DB2 Util Suite. Without IPLA Sub-Cap you have already
paid for install MSUs on the machines where DB2 runs. With the sub-cap IPLA
you only need to pay for the MSUs of DB2's own usage. It may growth with
your upgrade, but it is not until the DB2 MSUs are higher than your
previously installed capacity that you'll need more DB2 Util Suite Value Units. 

I recently completed just this analysis for a client. They have 840 Value
Units for their installed capacity of 3,848 MSUs spread across 6 machines.
The DB2 MSUs from the SCRT reports are 3,253 MSUs. That converts to a
requirement of 695 Value Units. So when they next upgrade, they already have
paid for 840-695=145 Value Units they do not need when they are sub-capacity
IPLA. Those value units will be used as DB2 continues to grow with increased
usage and growth in the business.  

You need to do a similar analysis for each product (or you could license a
tool  that reports this status with each set month's analysis.) 

When this was announced in August 10, 2004 your IBM may have told you it was
not important to you. That was probably true, you already the value units
necessary for your installed capacity. But you need to evaluate sub-cap IPLA
with each upgrade of any machine. 

Finally, Value Units belong to the enterprise, not to the site or to a
particular pricing plex. You can "transfer" them within your company. If the
data center in New York has excess value units and you need them at the
Chicago data center you can transfer them. 

Best regards,

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z/Expo questions

[Al Sherkow]

Lionel --

I've been attending z/Expo (now System z Expo) as a speaker since 1999. I
don't believe there is an online tool, though they have very good printed
agenda when you arrive onsite. There are not as many subject areas as at
Share: z/OS, z/VM, z/VSE and Linux. Within these areas sessions are
categorized into 12 groups such as G: General System z Technology, Z: z/OS
System Software and Parallel Sysplex, P: WLM and z/OS Performance, W: z/OS
Transactions, DB and Networking, L: Linux on System z, etc. 

The 'big' difference is that some of the very popular sessions are repeated
during the week. For example Kathy Walsh from the Washington System Center
is presenting "WSC z/OS Performance Hot Topics" on Wednesday at 4:10 and
Thursday at 2:35. Peter Enrico is presenting "WLM - Revisiting Goals in
2008" on Wed at 10:15 and Thur at 1pm. 

Sessions are 1h15m long. I add pages to my sessions and can cover more in
the session than at Share. 

There usually is (and I'm told there will be) a wireless connection
available throughout the conference area. I don't believe the sessions are
online, but you'll receive a CD of the entire conference which I often view
and take notes on during sessions rather than the paper handouts. They are
in PDF format, so to take notes on the files, you need some additional
software. I use Adobe Acrobat, but they may be other tools.  

Al Sherkow
I/S Management Strategies, Ltd.
+1 414-332-3062
Consulting Expertise on Capacity Planning, Performance Tuning, WLC, SW
Pricing, LPARs, IRD and LCS Software

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Upgrading the processor

[Al Sherkow]

Frank --

What are you doing for Disaster Recovery now? With current HW and SW you'll
be able to provide for DR. 

With old applications you were planning to move off you may have problems
getting zNALC pricing. zNALC requires non-legacy workloads or JAVA base
applications.  See IBM's US announcement letter 207-006 from 9Jan2007 and <
http://ibm.com/zseries/swprice/znalc.html > for additional information.

Evaluate your software pricing options. If sub-capacity Workload License
Charges will save money, then each month without sub-capacity charges will
cost you money.

Good luck! 

Al

-- 
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Capacity

[Al Sherkow]

Ed --

An LPAR may have an individual defined capacity and be part of a Group with
a Group Capacity Limit. The LPAR defined capacity can keep an LPAR from
having an 4-hour rolling average that is too high. The group capacity limit
keeps the simultaneous 4-hour rolling average of the group from being too high. 

Shane --

Sorry you missed the seminar in Sydney. It went very well. The next is
likely in Germany during November.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Group Capacity

[Al Sherkow]

The LPARs do not need to be in the same sysplex. This is one of the very
interesting features of the Group Capacity Limit. Sites have asked for this
for a very long, perhaps back to the beginning of Workload License Charges.
You can now set this parameter for all LPARs, and achieve for example: "The
Maximum Simultaneous 4-Hour Rolling Average MSUs That Are the Basis of SW
Charges Should Never Be More Than a User Specified Number". Something like
85% of the Machine.

z/OS (WLM) and PR/SM will manage the Group Capacity for you. You can Also
Have Multiple Groups regardless of Sysplex. Now you could limit the Maximum
Simultaneous 4-Hour Rolling Average MSUs from the Testing LPARs to Be Less
Than a User-Specified Number.

I added 30+ pages to my WLC seminar to cover Group Capacity limits. Even
more will be added when I actually have some data from a site using this new
capability.

To use this you need z/OS 1.8, and z9BC, z9EC at the appropriate hardware
microcode level. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: SCRT and IMS V7 question

[Al Sherkow]

In addition to Tom's suggestions you could look in the SMF30 data, the
measured usage segments to see which jobs/stcs are still using IMS. (In MXG
this would be the SMF30MU table.) The SMF89 will show you that IMS is
running, but not which addresses are using IMS. SMF30 will have the detail. 

I suspect you may still have a started running in support of IMS. I am not
an IMS expert, so I don't know where to look on the active system. 

(Thanks also to Tom for recommending LCS to help you manage, monitor, and
audit SCRT). 

Al 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: PR/SM -- WLM Capping

[Al Sherkow]

Rick --

You're really off here. PR/SM does not suspend logical processors for
"minutes to hours at a time". As Jim Mulder wrote Wednesday, it just does
not work that way. The suspension of the processor is a short duration. The
time-slicing generally dynamic and all the logical engines get an equal
share of an LPAR's allocation. This is not new since your 2064 it has always
worked this way (back to the 3090 where PR/SM was introduced).

The premise of your question is wrong, so it's hard to answer the question. 

You attributed some past performance problems being due to this issue, but
that is not correct either. There is or was another cause of your
performance problems.

Regards,

Al

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: zAAP Eligible Work

[Al Sherkow]

I believe it is JAVA applications that are using the IBM JVM (Java Virtual
Machine). If your ISV uses a "different" JVM then I believe it will not use
the zAAP engine. 

To accumulate data at the address space and service class period level on
how much zAAP and zIIP resources might be consumed if they were configured,
you use a new control specifiable in the SYS1.PARMLIB IEAOPTxx:
PROJECTCPU=YES. This will let you monitor if any applications will use the
specialty engines. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Workload License Charges & IRD Seminar in Sydney Australia

[Al Sherkow]

This post was approved by the list owner!

WORKLOAD LICENSE CHARGES SEMINAR in Conjunction with CMG Australia
May 22-24, 2007 in Sydney, Australia

This 3 day seminar will give you the essential knowledge you need to 
implement and leverage IBM's Workload License Charges, and PR/SM's 
Intelligent Resource Director. The seminar will help you answer these 
questions about WLC and IRD with respect to your site:


1. How can software costs be lowered with IBM's new WLC pricing model 
and its capabilities maximized?


2. How will LPAR Clusters and the Intelligent Resource Director (IRD), 
in conjunction with WLC, allow sites to meet Business Goals & Service 
Levels more effectively and more efficiently?


3. What are the Sub-Capacity Reporting Tool, the WLC Tool and the 
Sub-Capacity Planning Tool?


4. Should IPLA products be switched to zSeries sub-capacity pricing?

Registration is $1,900USD and discounts are available for early 
registration. Additional information and a syllabus of the seminar are 
available at: http://sherkow.com/publicseminarSydney.html



Al Sherkow

--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: License keys for ISV products(What alternatives are there?)

[Al Sherkow]

SCRT definitely only reports the IBM sub-capacity products for which it is
programmed. No ISV products are reported by SCRT today.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: WLM and WLC

[Al Sherkow]

It is a "statement of direction" that z/OS 1.8 will provide a Group Capacity
limit. This will be on z9EC and z9BC and later machines. (US letter: 
ENUS206-190.PDF). 

To change defined capacity on a set schedule your best bet is to use automation.

IRD can adjust an LPAR's weight to help satisfy goals, but this is not
related to Defined Capacity and billable MSUs.

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: zNALC in an LPAR and z/VSE V4 Sub-Capacity Pricing

[Al Sherkow]

z/OS.e and NALC are the same price. The primary difference being the
limitations of z/OS.e and that z/OS.e could be a sub-capacity pricing metric
while NALC was full capacity. 

Ed was correct that zNALC is priced lower than either once the billable MSUs
are above 45MSUs. The difference is not large, at 175 MSUs z/OS.e is $6,300
and z/OS with zNALC is $6,040 (USD). 

The big difference between NALC and zNALC is sub-capacity in an LPAR rather
than a dedicated machine. Also it seems more clear cut to me as to what is a
"Qualifying Workload" for zNALC. While the zNALC LPAR is to be dedicated to
the "Qualifying Workload" the database server used for the "Qualifying
Workload" can also be in the zNALC partition. This allows DB2 to be with
WebSphere if DB2 is the database server for a WebSphere "Qualifying Workload". 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: zNALC in an LPAR and z/VSE V4 Sub-Capacity Pricing

[Al Sherkow]

I believe zNALC will be usable for some inhouse written applications: From
the intro to the zNALC annoucement 207-006: "a qualified "new workload"
application (Qualified Application) such as Java™ language business
applications running under WebSphere® Application Server,"

and later in the detail of 2.(c): "(c) It is a Java language business
application running under WebSphere Application Server (or equivalent).
These do not include systems management tools."

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: z New Application License Charges (zNALC) Announced

[Al Sherkow]

My understanding is that LPAR name changes could be dynamic since the z990
was delivered. I've been advising people for years to have extra LPARs (yes
that is plural) defined to avoid the need for the POR, and the z990 allowed
sites to have extra LPARs defined and then to change the name to something
meaningful when they were ready to use those LPARs.

No one wants to be person to point out "Building that LPAR is a good idea,
but we need to shut down the whole machine to define it".

You have some time to rename your LPAR or define new LPARs because the
earliest date to collect data for SCRT is 2Apr2007. 

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

zNALC in an LPAR and z/VSE V4 Sub-Capacity Pricing

[Al Sherkow]

IBM announced zNALC today (http://ibm.com/systems/z/) though some of the 
links are not working. This allows NALC pricing for z/OS in an LPAR 
rather than a dedicated machine without some restrictions, but less than 
earlier NALC offerings


Also z/VSE V4.1 with sub-capacity pricing.

Al

--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Anyone converted from IFL to ICF??

[Al Sherkow]

On earlier zMachines these engines were all the same price in US$. On 
the z9BC announced in April 2006 the 'list price' is 95000USD for zIIP, 
zAAP, and IFL, but 125000USD for ICF.


Al

--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

LPAR Underload Concern

[Al Sherkow]

Hal --

It really depends on whether or not the defined capacity actually kicks in 
and begins to limit the LPAR with a soft-cap. If the 4 hour average goes 
above 4 MSUs the cap will be enforced. You can have much larger peaks 
without the cap, it depends on the duration of the peaks.


The real question is how many MSUs is that LPAR using on a regular and on a 
4 hour rolling average basis.


When the soft cap is on, the system will run as if it was out of CPU 
resource. Your WLM will determine which service classes are impacted based 
on your goals.


Best regards,

Al

--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062
Web: www.sherkow.com

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Subject: Re: SCRT questions

[Al Sherkow]

R.S. --

Yes, the LPARs must be running for the SMF data to be generated and 
collected. Sometimes, SMF data is lost, especially in 'sandbox' LPARs 
because managing this data did not matter until the machine starts to 
use sub-capacity pricing. It was certainly 'standard' practice to loose 
this SMF data. Now sysprogs everywhere need to learn to retain this 
data, and still some will be lost. These are 'sandboxes' after all!


Al

Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com




R.S. wrote the following on 7/5/2006 3:19 AM:

Al Sherkow wrote:


R.S. --

Removing LPARs is not allowed by the IBM Sub-capacity Ts&Cs. From the 
"Attachment for zSeries Workload License Charges" Z125-6516 (in the US)


   "2) collect, and retain for a period of not
   less than six months, the SMF data records
   for all LPARs, by Machine, required by the
   Sub-Capacity Reporting Tool for each
   Reporting Period;"

IBM's sub-capacity pricing is not an *LPAR* level decision, it is a 
*machine* level decision.


That's why I asked about it. However the above is not 100% true. It 
cannot be. Examples:
I have 7 OS LPARs, but 3 of them are not IPLed. In this case I 
*cannot* have SMF records from this LPAR. And SCRT does not include 
them in the report. However, if any of the LPARs was IPLed (i.e. for 2 
days) during the month, it will be included in the report. If you 
don't provide SMF records from such LPARs, the tool still says you 
have 100% records and the report is valid. Sometimes it is impossible 
to have records from 'sandbox' LPARs.
However it would be even better to include SMF records from one LPAR 
only. Better means the job would complete much faster.




--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Subject: Re: SCRT questions

[Al Sherkow]

R.S. --

Removing LPARs is not allowed by the IBM Sub-capacity Ts&Cs. From the 
"Attachment for zSeries Workload License Charges" Z125-6516 (in the US)


   "2) collect, and retain for a period of not
   less than six months, the SMF data records
   for all LPARs, by Machine, required by the
   Sub-Capacity Reporting Tool for each
   Reporting Period;"

IBM's sub-capacity pricing is not an *LPAR* level decision, it is a 
*machine* level decision.


If a machine is using sub-capacity pricing then the SMF70 data and the 
SMF89 data from all LPARs must be presented to the IBM's Sub-Capacity 
Reporting Tool (SCRT). If you are using Sub-Capacity, then z/OS is using 
sub-capacity and the LPARs must be reported. Also VM LPARs running z/OS 
guests and z/TPF.


The SCRT does report if LPARs are missing, as one MVS system's SMF70 
data has some information about the other LPARs.


If you strip off the SMF 70s and the SMF89s as suggested by Nico then I 
find the SCRT runs fast. The SMF70s and 89s are usually not too large. 
Most sites keep these in a DASD datasets.


Best regards,

Al

--
Al Sherkow, I/S Management Strategies, Ltd.
Consulting Expertise on Capacity Planning, Performance Tuning,
WLC, LPARs, IRD and LCS Software
Seminars on IBM SW Pricing, LPARs, and IRD
Voice: +1 414 332-3062 
Web: www.sherkow.com


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html