This is certainly of interest to computer geeks like us in general, but
I'm not sure why it would be of specific interest to mainframers.
According to this -
http://www-03.ibm.com/press/us/en/pressrelease/33233.wss - Watson is
powered by an IBM POWER7 server, or more than likely, a room of IBM
A detailed explanation would be nice. The way this article reads it's
the mainframe that's at fault. As though this wouldn't have happened
if the platform had been something other than the mainframe. However,
I'd bet that the upgrade was to the application system, not the
mainframe, and
RMF/SMF records timings, not percentages, so whatever analysis program
you are using is calculating the percentages. Typically that is a
percentage of the total machine/LPAR/engine, not based on the cap. That
is the percentage figure you'd see in the RMF reports. Of course, once
the LPAR use
Ron,
Are the libraries listed in the linklist on the two LPARs different? It
sounds to me as if you have a bogus IEFBR14 in the one linklist. I'd
check the libraries in the linklist on the LPAR where it is failing,
starting with the first library in the list, to verify that.
Tom Kelman
Yes, it sounds interesting. Did CA7 install its own version of IEFBR14?
As Gerhard said in another post, IEFBR14 as supplied doesn't have any
instructions requiring addressability, so it can't get an S0C4.
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
-Original Message-
From:
:
On Thu, 18 Nov 2010 00:42:13 -0600 Elardus Engelbrecht
elardus.engelbre...@sita.co.za wrote:
:Kelman, Tom wrote:
:We have it set up via the IEFU29 exit to submit a started task
to
dump the
:MAN file when a switch takes place. I do believe that this is
the
common
:process
Posted by Tom Kelman:
As long as the dump process is running properly SMF just switches
between the two large datasets, and the smaller dataset doesn't even get
used.
Reply from Ted MacNeil:
Since when?
In a lifetime away, we found that the smaller ones were gettin used, in
order.
Well, we
Ted,
I've never experienced that type of activity unless someone has turned
on something like a DB2 trace, and I have worked at a fairly large bank.
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
-Original Message-
From: IBM Mainframe Discussion List
Not only do you need to IPL (or at least restart SMF) to change the
CISIZE of the MAN files, the CISIZE of all the MAN files must be the
same, and the RECSIZE must be the same as the CISIZE. See pages 2-13
and 2-14 in the z/OS 1.11 version of the MVS System Management
Facilities manual for this
It also states that the CISIZE must be equal to the physical record
size.
Selecting the SMF Data Set Control Interval: The control-interval (CI)
size of SMF data sets can range from 0.5K (512 bytes) to 26K (26624
bytes) in size, with certain restrictions. The user specifies the CI
size of the SMF
Radoslaw,
What is your reasoning behind using many MAN datasets? The system only
records to one dataset at a time, unlike paging datasets which are all being
used. I've found that 2 large datasets and one smaller emergency dataset are
sufficient. As long as the dump process is running
Cross posted to IBM-Main and MXG-L.
We are looking at upgrading from DB2 v8 to DB2 v10 next year, and my
management has asked if DB2 v10 will end up moving any more processing
to our zIIP engine. By the time we upgrade to DB2 v10 we will also be
on z/OS 1.11.
Does anyone have any
Helen,
First, I would suggest you send an email to lists...@bama.ua.edu with
the message listserv refcard in the body of the email. This will get
you a list of all the commands you can submit to control your session.
For your specific question send the following to lists...@bama.ua.edu
Cobe,
Another thing is that as someone posted it looks like your trying to use
the hard cap by setting a weight of 160 with a dummy LPAR weighted at
10. That gives the real LPAR about 94% of the machine (160/(160+10)).
You then talk about wanting to set it to 24 out of 26 MSUs. That comes
to
That's true, but you'll dump less often.
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
-Original Message-
From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On
Behalf Of Ted MacNEIL
Sent: Tuesday, November 02, 2010 3:27 PM
To: IBM-MAIN@bama.ua.edu
Subject: Re:
Cobe,
You do have me a little confused when you talk about capping at 24 MSUs
and setting the WEIGHT (which I assume is the LPAR weighting factor) to
160. These are different. Also, the LPAR weighting factor is just a
relative number. It does not relate to MIPS.
As far as the LPAR going over
On my production system during prime shift (not quiet) it took about
35 seconds.
09:51:02.00 $TCK 0290 I SMF
09:51:36.49 INTERNAL 0290 START DUMPSMF,DSNAME=SYS1.MAN1
09:51:36.50 0281 IEE360I SMF NOW RECORDING ON SYS1.MAN2 ON
SYSC02 TIME=09.51.36
09:51:37.31 STC22057
If you are running a large CICS shop, and you have CICS TS v3 or higher,
make sure you have compression of the SMF 110 records turned on. That
will greatly reduce the amount of data put to the MAN files. Of course
your post processors will have to be able to decompress the records. I
know that
Interesting, but I wonder how much it relates to our reality.
From the web site about the program.
Outsourced is NBC's new workplace comedy series centered around a
catalog-based company, Mid America Novelties, that sells American
novelty goods including whoopee cushions, foam fingers and
His problem is certainly in ESQA. Based on the OMEGAMON screen he
showed he only has a total of 452K CSA. The SQA(ESQA) overflow was 887
pages. That's over 3M of storage.
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
-Original Message-
From: IBM Mainframe Discussion List
About 3 or 4 years ago we replaced the Compuware products with the
Macro4 products and have been very happy with them. Macro4 is now a
division of Unicom. The Macro4 products might not be as robust as
Compuware's, but they have been fine for our purposes, and cost less.
Strobe - FreezeFrame
Bob,
It's not working that way for me. At the top of the page it says
Welcome Tom Kelman. So it looks like I'm signed in, but when I click
on the link to download the PDF it says I'm not signed in. Something's
screwed up.
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
Yea, I tried the user ID and password I use for all other accesses to
IBM and it won't accept it. When I try to reregister with my email
address it says it's already in use. What gives?
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
-Original Message-
From: IBM Mainframe
That's my feeling also. In looking at the timeline on the Virginia
Government web site, it looks like it took about 24 hours to do the full
maintenance and repair of the EMC box. It was up by the morning of the
next day. However, it took a week to get the data completely recovered.
What took so
Personally I loved the way the article talks about silicon and germanium
as though they are brand new, just discovered materials. I think I
learned about silicon (atomic number 14) and germanium (atomic number
32) when I was in high school, and that's a long time ago.
Tom Kelman
Capacity
Well, I don't live there now, but I grew up in the Maryland part of the
Delmarva peninsula about 25 miles from Ocean City, MD., and I still have
friends there. Donna came through that area as a category 2 when I was
13. It had originally hit the Florida Keys and the southern tip of
Florida on
Cross posted to IBMMAIN and the MXG listserv:
I'm the mainframe performance analyst at my company. We have started
running our first intensive USS process. It is a package called Custom
Statement Format, and our application programmers have written a front
end to it. The front end is a UNIX
-
From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On
Behalf Of David Waldman
Sent: Friday, August 27, 2010 8:29 AM
To: IBM-MAIN@bama.ua.edu
Subject: Re: USS Processing - When are children spawned
On Fri, 27 Aug 2010 07:55:13 -0500, Kelman, Tom
thomas.kel...@commercebank.com wrote
...@bama.ua.edu] On Behalf Of
Kelman, Tom [thomas.kel...@commercebank.com]
Sent: Friday, August 27, 2010 5:55 AM
To: IBM-MAIN@bama.ua.edu
Subject: USS Processing - When are children spawned
I'm the mainframe performance analyst at my company. We have started
running our first intensive USS process
According to Cheryl Watson's CPU Charts the 2098-O02 is 359 MIPS as
opposed to 319 MIPS for the 2066-002. That's about a 12% improvement.
I'm not sure what you mean by 10 points but I would expect a CPU
utilization drop in the range of 10 to 12%.
Someone mentioned the technology MSU dividend. I
That is correct. What we see in out MXG JOBS dataset is the following
JOB
OMVS
STC
TSU
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
-Original Message-
From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On
Behalf Of Charles Mills
Sent: Wednesday, August 11,
I'm also curious as to why you would want to filter out the recording of
SMF records for certain datasets. SMF records are useful for a variety
of analysis tasks, not the least of which would be an audit trail in
case something happened to the dataset, or someone who shouldn't
accesses it.
One type of screw head that I haven't seen mentioned here is the torx,
or hexalobular, head. While the Philips screw was designed to cam out
of the screw to prevent overtightening, the torx screw, like the
Robertson screw, was designed to not cam out. It came about as better
torque-limiting
Well, this article is certainly a slam at Neon with this statement.
Neon Software is being a smart-alek by offering an IMS-only version of zPrime
for $1, and it is a shameless means of growing what is still a pretty small
installed base for the zPrime product.
And I love what they say at the
So, what do we in the United States of America call ourselves. We've
been called Americans, Americanos, or some other variation probably
every since we became a country. However, I was in Bolivia one year and
when I said to one of the Bolivians that I was an American he said You
know that we're
There are a couple of books available through Amazon that might be good
for an introduction. They are
Introduction to the New Mainframe: z/OS Basics by IBM Redbooks
z/OS (MVS) Primer by David Shelby Kirk
The second one has the Look Inside capability, and I checked the table
of contents. It
Charles, that's interesting. I worked in shops in the Southeast,
specifically Atlanta, for almost 30 years and there is was always kicks.
It's when I moved to the Midwest in 2005 that I got into a shop where
they said see-eye-see-ess, and the first time I used the term kicks they
didn't know what
I believe that the KICKS you're talking about is designed to run on
Hercules an MVS/zOS emulator that will run on Linux, various version of
UNIX, Windows, or Mac OS.
http://home.gci.net/~mike-noel/KICKS/
http://www.hercules-390.org/
Tom Kelman
Capacity Planning
Commerce Bank, Kansas City
Well, for IBM to be able to run Windows on the x86 blades I would think
that first there would have to be some sort of contractual agreement
between IBM and Microsoft to license the Windows OS. That might be more
difficult to accomplish that the technical aspects of the hardware
itself.
Tom
Of
Kirk Wolf
Sent: Thursday, July 22, 2010 10:08 AM
To: IBM-MAIN@bama.ua.edu
Subject: Re: System x and Power 7?
Tom,
How would you see it to be different from a regular System x BladeCenter?
Kirk Wolf
Dovetailed Technologies
http://dovetail.com
On Thu, Jul 22, 2010 at 9:45 AM, Kelman, Tom
thomas.kel
The TSS I know about is also known as TopSecret. It is a Computer
Associates product. You should be able to get the information on the
latest release here - http://www.ca.com/us/products/product.aspx?id=141.
CA also has the product ACF2. Both TopSecret and ACF2 are products that
replace RACF.
Since you say there is no money for a replacement, this will probably
not work. However, there is a SAS look alike product available that is
much less expensive. It even processes SAS statements. I doesn't do
everything SAS does yet, but it does have much of what's needed
including the
I would say that at a minimum you would want to be on separate power
substations. Your power company should be able to provide this service,
although they might charge extra for it. Most large IT organizations
feel it's worth it.
However, because of the interconnectivity in the U.S. and
Kees,
I do agree with you. See my post on massive power outages that have
occurred around the world. The only save plan is to have a backup for
the backup. You need to have a DR site on a separate power sub-station
from your main site along with a UPS and generator backup. Having
alternate,
I'm assuming that the electricity power cut, or whatever is happening,
is scheduled. You know ahead of time that it is going to happen.
If the power to your data center is going to be shut down for some
reason wouldn't you want to have a controlled power down of the complete
data center? I know
I was in a shop in the early 1980s where we had a very bad experience of the
power being cut by the use of the emergency switch. There was a building
maintenance man in the machine room. When he went to leave instead of pressing
the button to activate the door to the mantrap, he pressed the
I don't know why it wouldn't be getting the 21st, 22nd, etc. extent, but
why not allocate several EPILOG Datastores with no secondary extents.
We don't use EPILOG in my current job, but when I was using the EDS in a
previous job I had it set up that way. I had allocated 7 Epilog
Datastores with
I was wondering if anyone has tried to save a list of name to copy when
the SCRT reports are sent to IBM. There is a spot where you can enter
email addresses to receive a copy of the confirmation. I have to enter
7 addresses every time. Four of them are to managers in my company and
three more
We do collect Type 99 records. Do you collect the SMF 100-102 records
(DB2) and or the SMF 110 records (CICS) records? They have a lot more
volume in number of bytes recorded than the Type 99 records. Here is
the record count and sizes we collected yesterday. We are a small shop,
but the
This is a definite it depends not only as to what is running on the
LPAR, but also how one determines minimum. We have 3 LPARs defined on
a 4 engine z10BC of 1026 MIPS, 127 MSUs. Our production LPAR has 4 LPs,
so it could use all 127 MSUs if going full out. Our development LPAR
and our system
That is interesting, but I thought another part of the article was more
interesting. On the second page there is a heading Governments Invest
in Modernization with the following statement.
The timing of this research is intriguing; it comes on the heels of
reports that the U.S. Secret Service
, May 25, 2010 12:31 PM
To: IBM-MAIN@bama.ua.edu
Subject: Re: z/Vendor Watch: zNext or z11? Either Way, It's Coming
Soon!
-Original Message-
From: IBM Mainframe Discussion List
[mailto:ibm-m...@bama.ua.edu] On Behalf Of Kelman, Tom
Sent: Tuesday, May 25, 2010 12:23 PM
To: IBM
I think that Neon's problem with zPrime is that they are not using the
standard API. My understanding is that the standard API basically looks
for enclave work to be qualified for zIIP processing. Neon is moving
work other than enclaves to the zIIP. If BMC is using the API, then
there is no
OS/2 probably didn't have the security issues of Windows because (1) not
as many people used it, and (2) it wasn't around long enough for the
hackers to really get going on it.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816) 760-7632
-Original Message-
From:
Timothy,
I saw your post talking about having an A01 warm standby with Z03 CBU.
That's similar to what my management is thinking, but without the warm
standby. We have EMC storage and it we do synchronous mirroring from
our main site to our DR site. Since that is all done via the EMC
control
My management wants to benchmark the IPL process to see how long it will
take at various levels of MIPS, MSUs, or whatever measurement criteria
can be used. The purpose of this is to determine what the smallest z10
is we can contract for our DR site that will IPL in a reasonable time
frame. The
Bill,
My company has looked at z/Prime. Technically it appears to be a slick
product. However, there are outstanding lawsuits between Neon Software
and IBM. You can go to www.neon.com and see the information about the
lawsuits from Neon's perspective, or just Google neon zPrime ibm
lawsuit to
It is true that back in the good old days companies would have
internal training to teach programming skills. My first job after
college and the military was with a bank as an application programmer.
Back then they coded everything in IBM Assembler because it was more
efficient than COBOL. Up
, April 22, 2010 5:47 AM
To: IBM-MAIN@bama.ua.edu
Subject: Re: COBOL - no longer being taught - is a problem
On Wed, 21 Apr 2010 09:35:00 -0500, Kelman, Tom
thomas.kel...@commercebank.com wrote:
John,
That WikiPedia article also states that DMSII was created by
Burroughs
(later UniSys
That would be hard to say without really studying the current
environment. I would imagine that the current Unisys environment is as
far from what they are running on as would be System z using DB2 and
Cobol. However, since they are on an old Unisys/Burroughs environment
they might be able to
John,
That WikiPedia article also states that DMSII was created by Burroughs
(later UniSys) as a database to run on its processors. Does that mean
they are still running UniSys machines. If so they have problems over
and above COBOL not being taught. It sounds like a typical government
This is all very interesting. My first real programming job was as a
coop student at a DuPont nylon manufacturing plant in Seaford, Delaware.
I was in the Works Engineering Department which was responsible for the
onsite power plant, the electricians, the electronics shop, and
engineering
If you're using standard I/O processing in your application code, you
shouldn't have a problem. We just converted 3 files that are used in
our CICS systems to VSAM Extended and didn't need to change any
application code.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816)
I agree with John. Since I, as the performance analyst/capacity planner
here, am responsible for sub-capacity pricing processes, I can easily
install SCRT into libraries that I control. Since it has no affect on
external customers, I don't have to go through heavy duty change
control. If it
I just received an email that comes from nets...@us.ibm.com, or so the
addressing says, which says IBM found my business information at
www.jigsaw.com and based on that they will be sending me insights,
information and offers. The URL www.jigsaw.com is blocked by my
company as is the opt out
That's what I figured. Thanks.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816) 760-7632
-Original Message-
From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On
Behalf Of Staller, Allan
Sent: Friday, April 09, 2010 7:50 AM
To:
(816) 760-7632
-Original Message-
From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On
Behalf Of R.S.
Sent: Wednesday, April 07, 2010 10:43 AM
To: IBM-MAIN@bama.ua.edu
Subject: Re: JCL QUESTION
Kelman, Tom pisze:
Just a warning about using SPACE=(TRK,0). We had
You see the SMS allocation because if the dataset doesn't exist the JCL
you've specified will allocate it as a new dataset and then delete it.
If it does exist it will be allocated at and old (mod) dataset and then
deleted. In either case you should see this message later in the
JESYSMSG.
Paul,
I was surprised. I thought the type of space (TRK, CYL, etc.) was
required. I just tried allocating a dataset using SPACE=(1,1), and it
does allocate 1 track. However, personally I still prefer putting TRK
in the SPACE parameter. It's called documentation, and there are
probably others,
I have written several WLM service policies, but I've never written code using
the WLM macros. So I looked in the MVS Programming WLM Service Guide to see
what it was all about. From what I gather, this particular macro is used to
gather WLM information so that a scheduling service can make
Just a warning about using SPACE=(TRK,0). We had a job that used that and it
all of a sudden started getting abends with a message that there was no space
defined. Our storage folks had just set up SMS for VSAM Extended for that
specific group of datasets. It appears that once you do that it
Yes, OMEGAMON also has the ability to monitor the 4HRA when the LPAR is
soft capped.
In the SMF Type 70-1 record, the variable SMF70MSU has the defined
capacity for the LPAR and SMF70VPF is a flag where one of the bits
indicates the partition capping is enabled. However, I don't know if
that
z/XPF is a pretty good product. I believe it is the one a previous post
mentioned that user the trace tables. I saw the demo at CMG 2009. The
only problem for my shop is that it doesn't get inside CICS yet, but
they said they're working on that.
Tom Kelman
Enterprise Capacity Planner
Commerce
I was just reading a presentation made by Rich Olcott at the August 2008
SHARE that might relate to what Hal is referring to. In it Rich is
discussing the unattainable goal. He says that after a time WLM will
realize that the goal can't be reached and will stop trying for a while.
So basically
Always check your velocity goals when changing hardware. Those can be
affected by the differences in hardware speed. You response time goals
should be OK.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816) 760-7632
-Original Message-
From: IBM Mainframe
The big problem is that no one knows how to read, or write, anymore. I
was at a well known, excellent engineering school not long ago. I
happened to pass by a bulletin board that held postings put up by
faculty and students to announce things that were happening around the
campus. The spelling
One thing to be aware of is that once you turn on zAAP on zIIP you no
longer get the zAAP information in the SMF records (i.e. zAAP elligible,
etc.). It is now all considered zIIP work and shows up in those
buckets.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816)
I have read through several responses to this, and I agree with all of
them. Performance wise, having less LPARs is probably better. Our shop
is small and we have only on CEC with three LPARs (production,
development, and sysprog sandbox). One purpose for the three is to be
able to test new
Yes, it is not a good idea to assign all of your CPs (physical
processors) to all of your LPARs. What you are doing is having all the
LPARs competing for the CPs on an equal basis. The weights take effect
only when the total CEC is 100% busy or at least very close to it.
Also, the recommendation
I am in the process of completely redesigning our WLM policy, so I'm
going through pretty much the same as you. Although, I do have some
experience in designing one shortly after WLM appeared on the scene.
What I'm trying to get a handle on is new functionality and subsystems
that have been
Neon has posted a response to IBM's countersuit on their web site here -
http://www.neon.com/neon/countersuit.shtm. These are several documents
that have been filed in the US District Court, Western District of
Texas, for all you that are interested in that.
This could get real interesting.
When setting up CICS you would normally place the CICS regions into a
velocity service class as you have done. In determining the velocity of
a service class, WLM samples the service class periodically and
evaluates the velocity as (no. of using samples)/(no. of using samples +
no. of waiting
Steve,
I've been thinking about this, and I also saw Steve Thompson's post
about it. I've been using Windows for years. I'm currently on
Windows/XP thinking about going to version 7. I've also thought about
getting a MAC because I do a lot of Photoshop editting and I've been
told the MAC is
I also agree that you should keep it simple. Also, remember that the
importance level determines just what it says and the velocity goal
should be set according the type of work running in the SC. Don't set
the goal based on the importance of the work. Assuming all 6 service
classes run similar
Allan,
I'm interested in your statement about DB2 being managed at the
transaction level (usually). As far as I know DB2 is not transaction
managed the way CICS or IMS is. DB2/DDF does do its processing via
enclaves, but that is different from the type of transaction processing
that is done for
Thank you to all those who've replied concerning this. We probably are
going to turn it on and give it a try. Does anyone have a good idea of
what to measure before and after we flip the switch on HiperDispatch to
be able to prove whether or not it helped at all? Any ideas are
welcome. Also,
Cross posted to the MXG listserv.
I'm trying to verify and consolidate what I feel I know about
hiper-dispatch. We are running a z10-BC under z/OS 1.9. Our processor
has 4 engines and 3 LPARS (production - 4 LPs, development - 2 LPs, and
system programmer sandbox - 2 LPs0). We currently are
-Original Message-
From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On
Behalf Of R.S.
Sent: Tuesday, December 01, 2009 3:44 AM
To: IBM-MAIN@bama.ua.edu
Subject: Re: Now is time for banks to replace core system according to
Accenture
Kelman, Tom pisze:
Accenture is saying
According to the SMF when a Z EOD command is done the MANx files are
switched.
When the operator issues either the HALT EOD or the SWITCH SMF command,
the following actions occur: v A type 19 record is created for each
online direct access device (if a type 19 record was specified) v For a
SWITCH
Accenture is saying that now might be the time for banks to replace
their core systems and retool to new technology. They don't actually
say it, but it sounds to me like their saying that big banks should get
off the mainframe. comments?
Yes, some vendor software does have a grace period so it will work
during a cut over/upgrade or in the case of a disaster recovery test or
the real thing - others don't. The problem is that the really critical
pieces of software usually don't.
Why do you want your vendors sleeping during the
That's what I was thinking. A system this critical and there is no
backup and no failover for 7*24 uptime. Just think - this is one of the
systems that controls our airline flights. I was talking with someone
that I work with who is actually an ex-employee of one of the major
airlines. The
We went all virtual tape about 6 months ago. We have kept three 3490
drives because we do deal with some customers that still require tapes
be sent. The drives are standalone and require that the tape be mounted
by a human operator the old fashion way.
Tom Kelman
Enterprise Capacity Planner
As a follow on to my previous reply to Jim's post. We mirror all the
data that we need for disaster recovery purposes to our DR site about 30
miles away.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816) 760-7632
-Original Message-
From: IBM Mainframe
We do the same. We have EMC disk and virtual tape systems which are
fully mirrored to our DR site. The processor at or DR site is the same
model as the one in our main shop but with only 1 engine available. We
do our DR testing using that one engine and will use capacity on demand
to bring up
You said that the module was linked reentrant, but are you sure it is
truly reentrant? Are you sure it doesn't have code that modifies
storage making it a non-reentrant module?
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas City
(816) 760-7632
-Original Message-
From:
The 1401 (actually a 1410) using the Autocoder language was the first
IBM computer I was involved with. I was a college student and had a
co-op job with DuPont. About a year after I started working with them I
returned after my school quarter and they had replaced the 1410 for an
IBM 360 and an
__
Thomas Berg Specialist IT-U SWEDBANK
-Ursprungligt meddelande-
Från: IBM Mainframe Discussion List
[mailto:ibm-m...@bama.ua.edu] För Kelman, Tom
Skickat: den 17 november 2009 14:54
Till: IBM-MAIN@bama.ua.edu
Ämne: Re: A big contributor to S/360
But instead of downgrading why not get the z10BC? It has a better MSU
to power ratio so you get the savings in software costs without the
decrease in power. Besides, I believe the zIIP and zAAP engines are
also less expensive. That's what we did. We went from a 2096-U04
without a zIIP to a
An XML engine was added to SAS in version 8.2 and has been improved
since then. You should be able to write a SAS program, or have someone
write it, that will pull the data you want from the MXG database and
output it in XML format.
Tom Kelman
Enterprise Capacity Planner
Commerce Bank of Kansas
1 - 100 of 377 matches
Mail list logo
