Hi
MIPS topic again:
One of our customer has reported that our (long runing) server
application is using 500 MIPS /day.
How to understand this, and is there any standard tool to generate a
report like this(i.e how many MIPS has used )
--
Miklos Szigetvari
Development Team
ISIS Information
Steve,
This particular kind of thing is sometimes seen where a program is
NON-LE compliant and gets called by LE compliant code. R12 is not what
LE want's it to be (and x'000c' is NOT what LE wants) and so it
augers in and makes a crater.
I also currently assume that I have to look in
Miklos Szigetvari wrote:
Hi
MIPS topic again:
One of our customer has reported that our (long runing) server
application is using 500 MIPS /day.
How to understand this, and is there any standard tool to generate a
report like this(i.e how many MIPS has used )
I understand it as a
instructions/time/time? That's acceleration.
Either they've got a very serious problem or its not being reported properly.
i
-- Original Message --
Received: Thu, 14 Aug 2008 08:13:00 AM BST
From: Miklos Szigetvari [EMAIL PROTECTED]
To: IBM-MAIN@BAMA.UA.EDU
Subject: MIPS /day
Hi
One of our customer has reported that our (long runing) server application
is using 500 MIPS /day.
How to understand this, and is there any standard tool to generate a report
like this(i.e how many MIPS has used )
IBM doesn't use MIPS, so if you feel MIPS is a valid measurement you will have
Mips/Day? Is this a short way to tell A mips a day keep the upgrade
away?\\Itschak
On 8/14/08, Ian S. Worthington [EMAIL PROTECTED] wrote:
instructions/time/time? That's acceleration.
Either they've got a very serious problem or its not being reported
properly.
i
-- Original Message
On Wed, 13 Aug 2008 08:40:20 -0700, Craig Bakken
[EMAIL PROTECTED] wrote:
Are there any caveats to moving the ownership of the filesystem while the
file system is in use or is OMVS pretty well behaved? When we do rolling IPLs
the TWS/TWSE2E tasks are moved to an active system. Should we move
Here is a sample of what we use. Feel free to contact me offline if you
have questions.
Jon
STR14,VSMREGSB SET REGION SIZE TO REQUESTED SIZE
ALR14,REGADJ ADJUST GETMAIN LIMIT (ADD 512K).DMH
C R14,DEFBELOW COMPARE TO DEFBELOW
IBM doesn't use MIPS
Actually, they do.
I have an old slide chart from IBM Canada, where a 3090-200E is 31.5
'processing units'.
-
Too busy driving to stop for gas!
--
For IBM-MAIN subscribe / signoff / archive access
Did you put TCP tcpname in parenthes?
ping 127.0.0.1 (TCP TCPIP
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Mohd
Shahrifuddin
Sent: Wednesday, August 13, 2008 10:40 PM
To: IBM-MAIN@BAMA.UA.EDU
Subject: TSO PING command
Dear all,
Try
Your customer is using ManagerSpeak. They are using the term MIPS in a
different context from the computer professionals on this list. You need to
be a bit of a translator from ManagerSpeak to ComputerProfessional.
Your customer is saying that the long running application is taking a large
There is a third option, but could be the riskiest. Verify that the
libraries in the JOBLIB only contain modules from TRUSTED sources, and add
them to the APF list.
Wayne Driscoll
Product Developer
NOTE: All opinions are strictly my own.
-Original Message-
From: IBM Mainframe
Eric Bielefeld wrote:
I finally got a new job, starting Monday Aug. 18. As I don't know how my
employer will react to posting on IBM-Main, I'll wait to tell you who after I
start. I'll be working in St. Louis.
Anybody in St. Louis from IBM-Main? Contact me off list, or call my cell phone
Never mind that, I saw following message about the JOBLIB containing
application libraries.
Wayne Driscoll
Product Developer
NOTE: All opinions are strictly my own.
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
Of Wayne Driscoll
Sent:
Getting the 'star' treatment is the all new 3390 Model A introduced in
z/os 10. This is a logical DASD volume with a slight increase in the
architectural limit of 65,520 cylinders to 268,434,453 cylinders. Gulp.
Did he say 268 mega cylinders? Yup.
What is amusing is that many of us know and love
-Original Message-
From: IBM Mainframe Discussion List
[mailto:[EMAIL PROTECTED] On Behalf Of Hal Merritt
Sent: Thursday, August 14, 2008 10:18 AM
To: IBM-MAIN@BAMA.UA.EDU
Subject: Here we go again (was z/OS Hot Topics 19)
Getting the 'star' treatment is the all new 3390 Model A
So if you go to the new EAV - I wonder how many DR Sites are actually ready
for them? At Share I have not heard any information on that side of the
equation. I am sure companies with internal DR Sites will be fine. But
what about those companies that contract out for this service. Do you think
Yup- planning is well underway at my shop. In fact, we have even
selected the volume serial for the first one: C:
Just kidding.
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
Behalf Of Lizette Koehler
Sent: Thursday, August 14, 2008 10:29 AM
To:
-Original Message-
From: IBM Mainframe Discussion List
[mailto:[EMAIL PROTECTED] On Behalf Of Lizette Koehler
Sent: Thursday, August 14, 2008 10:29 AM
To: IBM-MAIN@BAMA.UA.EDU
Subject: Re: Here we go again (was z/OS Hot Topics 19)
So if you go to the new EAV - I wonder how many
McKown, John wrote:
Getting the 'star' treatment is the all new 3390 Model A introduced in
z/os 10. This is a logical DASD volume with a slight increase in the
architectural limit of 65,520 cylinders to 268,434,453
cylinders. Gulp.
Did he say 268 mega cylinders? Yup.
Nah, call it the
I also had a same issue when my client wanted to know, how much horse
power was being used by his application, and no such standard tool I could
find out there.
Instead I used the following formula;
CPUTIME / (ACTUAL INTERVAL DURATION * NUMBER OF CPS) * TOTAL MIPS
In a message dated 8/14/2008 10:50:44 A.M. Central Daylight Time,
[EMAIL PROTECTED] writes:
I also had a same issue when my client wanted to know, how much horse
power was being used by his application, and no such standard tool I could
find out there.
Instead I used the following
Instead I used the following formula;
CPUTIME / (ACTUAL INTERVAL DURATION * NUMBER OF CPS) * TOTAL MIPS
Unfortunately, not all of the CPU consumed by a task is collected.
These days about 80-85% is reported.
Highly swappable work has a lower ratio.
So, you can only be relative to a given task
In a message dated 8/14/2008 10:54:21 A.M. Central Daylight Time,
[EMAIL PROTECTED] writes:
And this computation results in some number of foot-pounds per second? lol
More closer to stone furlongs per fortnight...tie the galvanometer to the
dynamometer and divide by Pi R**2Guess
-Original Message-
From: IBM Mainframe Discussion List On Behalf Of Edward Jaffe
McKown, John wrote:
Nah, call it the 3390EL for Excessively Large. I cannot imagine
backing that monster up. But, then, compared to my PC's 500
Gb drive???
500 GB is nothing compared to the
snip-
MIPS topic again:
One of our customer has reported that our (long runing) server
application is using 500 MIPS /day.How to understand this, and is there
any standard tool to generate a report like this(i.e how many MIPS has
..and then think...a disk failure on a one-pack sysplex. Scary!
John Compton
Phone Cork: +353 (0)21 231 4641;
Phone VOIP: 214-775-3641
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
Of Chase, John
Sent: 14 August 2008 17:25
To:
Hi, All,
In the z/OS 1.9 MVS Init 'n' Tuna manual I read:
Begin paste
SYSLIB statements must always appear before any LNKLST statements in
PROGxx. If you specify multiple PROG=xx members, define any SYSLIB
statements ahead of LNKLST statements. For example, if you specify
Ed,
The initial size limit of the EAV is still quite small compared to common
usage on other platforms and applications.
It is quite common for Open Systems to have just one to LUNs (volumes) per 8
disk parity group, so volume sizes of 100 to 500GB are quite common. Some
applications like film
John,
Not really. A disk failure - two disk failures actually - on an Array Group
of 3390-3 can affect over 300 volumes. I would much rather manage restoring
4-10 large volumes than 300+ small ones.
Ron
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED]
Ted,
I think a lot of that uncaptured time for addresses with a lot of DW and LW
swapping was mitigated when RCT time was added into CPUTM (using MXG
parlance).
Ron
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
Behalf Of Ted MacNEIL
Sent:
John,
SYSLIB statements must come before LNKLST entries. Much simpler IMO to
create a separate PROGxx member with just the SYSLIB statement(s).
Lot of sysprogs like to throw everything into one PROGxx member but I find it
easier and clearer to separate them. For example, PROGxA for APF
On Wed, 13 Aug 2008 09:58:19 -0500, Tom Marchant m42tom-
[EMAIL PROTECTED] wrote:
If all else fails, you could REPRO MERGECAT everything else to a new catalog
and delete the catalog.
If there are no other datasets beyond the valid qualifier level, you might
also
be able to REPRO MERGECAT just
We need to point to a new DNS server. Can this be done without stopping
TCPIP? Maybe through a modify command?
Jerry
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the
A while back I tried to do this with an obeyfile but received error
messages regarding invalid operand.
-Original Message-
From: Jerry Fuchs [mailto:snip]
Sent: Thursday, August 14, 2008 11:33 AM
To: IBM-MAIN@BAMA.UA.EDU
Subject: Changing the NSINTERADDR TCPIP parm dynamically.
We need
You can change/add the NSINTERADDR in your TCPIP.DATA file and issue F
RESOLVER,REFRESH.
Jeff Beech-Garwood
Systems Programmer
Mainframe Network Support
US Bank, EP-MN-02NS
651-962-3809
Jerry Fuchs [EMAIL PROTECTED]
Sent by: IBM Mainframe Discussion List IBM-MAIN@BAMA.UA.EDU
08/14/2008
I think the original point was more about logical volume failures (VVDS
deleted, VTOC or VTOCIX hosed). The larger the volume, the greater the
impact of a failure.
After time, I wonder what fragmentation would look like.
-Original Message-
From: Ron Hawkins [mailto:snip]
Sent:
Thanks. That's what I needed.
Jerry
Jeff Beech-Garwood [EMAIL PROTECTED]
Sent by: IBM Mainframe Discussion List IBM-MAIN@BAMA.UA.EDU
08/14/2008 02:42 PM
Please respond to
IBM Mainframe Discussion List IBM-MAIN@BAMA.UA.EDU
To
IBM-MAIN@BAMA.UA.EDU
cc
Subject
Re: Changing the NSINTERADDR
Some performance improvements in that process are included.
Also, a (extra cost?) 'hyper PAV' that uses a 'ucb' from a pool only for
the duration of the I/O. Talk about instant gratification :-)
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
On Thu, 14 Aug 2008 11:41:29 -0700, Jeff Beech-Garwood
[EMAIL PROTECTED] wrote:
We need to point to a new DNS server. Can this be done
without stopping TCPIP? Maybe through a modify command?
You can change/add the NSINTERADDR in your TCPIP.DATA file and
issue F RESOLVER,REFRESH.
...
If you
Ted,
I think a lot of that uncaptured time for addresses with a lot of DW and LW
swapping was mitigated when RCT time was added into CPUTM (using MXG parlance).
There is still a portion missing.
It used to cost about 800 instructions to measure an event (approximately).
So, some events are not
ok---so far so goodquestion on NODE() and the APPL()
I gather from manual I can have APPL(name1) NODE=2
APPL(name2)
NODE=2
APPL(name3)
NODE=2
if NODE(2)
Or, if you are paranoid, this is an effort by the customer to determine
if the vendor has both the technical knowledge and the integrity to come
back with a response to the effect This is a meaningless issue and we
don't want to waste your money dealing with it.
-Original Message-
From:
Situation came up on a client site where they received a very large file,
2.988 million records, and allocate the file size at (cyl,(2500,100)) with a
unit parameter of sysda,10; the file is received across 2 volumes in 20
extents totaling 4,331 cylinders. I am not concerned about the extents,
a z/OS batch job FTPing an MVS dataset to a non-MVS (perhaps windows)
server. The data contains a .. (two periods), which seems to be treated on
the receiving end as a CRLF (they are stripped out and the following
characters are on a new line justified left). My thought is specifying ASCII
in the
On Thu, 14 Aug 2008 17:22:35 -0400, William F Besnier wrote:
Situation came up on a client site where they received a very large file,
2.988 million records, and allocate the file size at (cyl,(2500,100)) with a
unit parameter of sysda,10; the file is received across 2 volumes in 20
extents
On Thu, 14 Aug 2008 16:37:12 -0500, Paul Peplinski
[EMAIL PROTECTED] wrote:
... The data contains a .. (two periods), which seems to be treated on
the receiving end as a CRLF (they are stripped out and the following
characters are on a new line justified left). ...
A little more info would be
What is the LRECL, RECFM, and BLKSIZE? 4331 cylinders is 64,965 tracks.
That averages to ~46 records per track. Is there some reason you think
this is unreasonable?
What is the percentage used of the 4,331 cylinders? What is the device
type the dataset is on?
If the initial allocation
We are trying to research what software is producing SMF subtype of 132?
We see these records being created every 1-2 minutes when a TSO user is log
on the system.
Any ideas?
--
For IBM-MAIN subscribe / signoff / archive access
I'm not sure I understand what you mean; if the default volume is a 3380 and
the allocation is satisfied on a 3390, a larger volume, how would the space
allocation be increased for the larger 3390.
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
Of Schwarz, Barry A
Sent: Thursday, August 14, 2008 6:44 PM
To: IBM-MAIN@BAMA.UA.EDU
Subject: Re: DASD Space Allocation
What is the LRECL, RECFM, and BLKSIZE? 4331 cylinders is 64,965 tracks.
That
Rogers,
you said subtype 132, do you mean type? if so, I see references to
NETSPY. If you really do mean subtype I saw references to type 102 (DB2)
with a suptype of 132.
Thanks,
Fletch
snip
We are trying to research what software is producing SMF subtype of 132?
We see these records
Please explain how 1831 cylinders requires 19 extents and that is consistent.
Maybe there is fragmentation on the secondaries?
When your allocation goes to secondary volumes, only the secondary extent is
used.
The primary is NOT remembered.
The primary could be broken up depending on the
On Thu, 14 Aug 2008 20:52:01 -0500, Rogers Laine [EMAIL PROTECTED] wrote:
We are trying to research what software is producing SMF subtype of 132?
We see these records being created every 1-2 minutes when a TSO user is log
on the system.
Any ideas?
Dump a few of these records and look for an
I will be out of the office starting 08/14/2008 and will not return until
08/18/2008.
I will be out of the office Friday, August 15. I will return on Monday,
August 18th. Thanks.
**
The information contained in this communication is confidential, private,
proprietary, or otherwise
William,
Assuming you do not have any software with Primary or Secondary extent
reduction, or secondary extent increase rules. I'm talking pure dataset
allocation without any allocation recovery.
So, if the Primary was allocated in four extents, that means you can have an
additional 12 extents
Ted,
Isn't RCT time a clock duration measured the same way as TCB and SRB time
and accumulated in the RCT Time bucket?
Ron
There is still a portion missing.
It used to cost about 800 instructions to measure an event
(approximately).
So, some events are not worth measuring.
-
Isn't RCT time a clock duration measured the same way as TCB and SRB time and
accumulated in the RCT Time bucket?
I honestly cannot remember.
Regardless, not all CPU consumed is measured.
-
Too busy driving to stop for gas!
--
58 matches
Mail list logo