does it have to be a gdg?
what about mod'ing onto the file then emptying it out when it's processed?
back it up with a gdg with that batch job at regular intervals suggestion.
From: Donnelly, John john.p.donne...@ti.com
To: IBM-MAIN@bama.ua.edu
Sent:
why not ask CA?
From: Pascoe, Raymond M raymond.pas...@highmark.com
To: IBM-MAIN@bama.ua.edu
Sent: Tuesday, May 8, 2012 11:06 AM
Subject: Re: National Vulnerability Database (NVD) Search for Mainframe
Vulnerabilities
Not sure if this forum is the
One more thing about keeping the rexx efficient.
I like to write structured code so that no code is executed unnecessarily, plus
it's easier to update.
I didn't run a test check the difference between compiled interpreted rexx,
but here's the cpu usage results from one of my (uncompiled)
when I wars responsible for application software 10 and/or 20 years ago, I did
my usual low tech, basic utilities solution. I'd do a 'listd hist' on the
member names and store the results. programs were compiled when they went from
dev to test, and copied from test to prod/staging. yes I
It really depends on what you're doing with REXX that's causing it to munch
CPU. I associate CPU usage with slow processing, and since I'm not the perf
tune guy, I really only care about how long I have to wait for my code to
execute to completion and do so without adversely affecting others.
if it's a 24x7 shop, with 1 lead/mgr per shift, that sounds about right for any
size unless workload is large.
--- On Fri, 2/10/12, Dan Gillis gillis...@gmail.com wrote:
From: Dan Gillis gillis...@gmail.com
Subject: Question on Job Scheduling
To: IBM-MAIN@bama.ua.edu
Date: Friday, February
well, whatever runs in production, you're going to need a copy of it in
testing.
usually means a lot of batch and then testing the onlines/applications' user
access.
I found it always helped to keep a testbed of data for batch/db's, then write
an automated batch job scheduler that edits all
]
On Behalf Of Cris Hernandez #9
Sent: Tuesday, January 24, 2012 12:38 PM
my sysprogs use WLM to wreck all kinds of havoc on my JCL
parameters and the otherwise normal (default) OS
functionality.
--
For IBM-MAIN subscribe
my sysprogs use WLM to wreck all kinds of havoc on my JCL parameters and the
otherwise normal (default) OS functionality.
--- On Tue, 1/24/12, Tim Brown tbr...@cenhud.com wrote:
From: Tim Brown tbr...@cenhud.com
Subject: change job classes for ones submitted via intrdr
To:
when I want to know what a job did, I read it's output using REXX, whether it's
the util/pgm output to a file or the JES output pulled off the output queue or
SAR. what is it you get from SDSF?
--- On Mon, 1/16/12, Al Chu al_chu...@optusnet.com.au wrote:
From: Al Chu
I suspect ITIL (cert org?) refers to mainframes by their proper nomenclature as
designated by the manufacturer and ditto it's OS. To me, mainframe is
synonymous with IBM for both manufacturer and OS, but Unisys uses the term
mainframe as well. No doubt people here will like to be exact with
On Thu, 15 Dec 2011 14:17:27 -0500,
Shmuel Metz (Seymour J.) wrote:
on 12/14/2011 at 05:21 PM, Cris Hernandez #9 said:
yeah, TRANSLATE works to change it after the user
hits enter, but I
want characters to show in caps as soon as the
character is typed,
How could ISPF do
yeah, that sounds like what I want to do, but to my knowledge there's no way to
get an spf panel display to read/replace a character as soon as it's typed. no
wonder web developers loath the mainframe... it isn't just a lack of pictures.
I'll check the links pubs ref'd in this stream (thanks
Subject: Re: Is there an SPF setting to turn CAPS ON like keyboard key?
To: IBM-MAIN@bama.ua.edu
Date: Wednesday, December 14, 2011, 8:37 PM
On Wed, 14 Dec 2011 17:21:11 -0800,
Cris Hernandez #9 wrote:
yeah, TRANSLATE works to change it after the user hits
enter, but I want characters to show
just wondering if there's a way to turn CAPS ON using an SPF setting so I can
set it in a REXX exec so typing occurs in uppercase without having the user
hitting the CAPS LOCK key.
--
For IBM-MAIN subscribe / signoff /
TRANSLATE in the
REXX exec to make it uppercase regardless of the input?
Sent from my iPad
On Dec 15, 2011, at 13:51, Cris Hernandez #9 hernandez...@yahoo.com
wrote:
just wondering if there's a way to turn CAPS ON using
an SPF setting so I can set it in a REXX exec so typing
occurs
any webpage will do if the report is properly formatted for the software used
to view it. security considerations are site specific. I prefer not to have
the overhead of maintaining a webpage. Instead, I prefer to format my reports
for simple text/wordpad viewing on the pc if no further data
IBM can't find people to staff their help center in Debuque Iowa??? 200 miles
NORTHwest of Chicago?? I got a call from an IBM staffer when they were
standing that place up. They were so desperate, they were going through their
files of former employees. Once was enough for me, 2nd time as
I'd start with standard IBM utility (iebupdte) to unload the pds to a
sequential file, send it, then use same utility to create a new pds (or update)
from the sequential file. if it gets screwed up in transmit, deal with it when
it happens, the messages in the utility output should indicate
this begs the question... why?
If you need to use the ID, why not just change the password?
only answer I can come up with is that the ID's password is embedded in a bunch
of stuff across applications and/or platforms.
if that's the case, then you got procedural issues to iron out with
Like these kind of stories...
I basically had a fit when I first heard we were losing our card readers since
I had no other way to run my batch print jobs. No one bothered to teach my how
to use FQ ahead of time. The year was 1980. We kept the key punch operators
for a few more years after
what error messages are issued when you attempt to logon?
any indication of the connection being made when you open your mainframe's IP?
need to see all that to have any idea of what happening.
--- On Fri, 8/5/11, saurabh khandelwal sourabhkhandelwal...@gmail.com wrote:
From: saurabh
When it comes to unix (or winders), I really don't mind being the village
idiot, because at least I know what I don't know, or at least I thought so
until I read this rant. A lot of my manuals refer to (what I call) mainframe
unix as OMVS and/or USS. Not being a seasoned uniq, all these
use browser and ftp instead of http?
--- On Mon, 8/1/11, Ed Gould ps2...@yahoo.com wrote:
From: Ed Gould ps2...@yahoo.com
Subject: FTP from IBM's book server
To: IBM-MAIN@bama.ua.edu
Date: Monday, August 1, 2011, 1:32 AM
John had said he was able to FTP and
download books from IBM's book
Gilmartin paulgboul...@aim.com wrote:
From: Paul Gilmartin paulgboul...@aim.com
Subject: Re: REXX code on JES2
To: IBM-MAIN@bama.ua.edu
Date: Saturday, July 30, 2011, 8:44 AM
On Fri, 29 Jul 2011 15:52:45 -0700,
Cris Hernandez #9 wrote:
what I usually do with my rexx programs is set one
thanks for pointing that out, I totally
forgot about those negative RC's and what uses them.
is it the LISTDSI ? EXECIO ?
I was mainly thinking about the TSO commands.
--- On Sat, 7/30/11, Ted MacNEIL eamacn...@yahoo.ca wrote:
From: Ted MacNEIL eamacn...@yahoo.ca
Subject: Re: REXX code on
:13 -0700,
Cris Hernandez #9 wrote:
First thing I would do is figure out which statement is
causing the error.
I use a simple
lc = RC
to capture the return code from all commands
calls (like ALLOC, EXECIO, FREE, etc.), then code:
Why capture RC? It's not going to escape.
IF lc 0
can't spoof a legit email address or spearphish an admin using a job ad, but
some companies are learning to be more careful about the amount of detail they
post.
I always find it kind of creepy when I get email from myself that I didn't
send...
--- On Fri, 7/29/11, Steve Conway
two more thoughts from the mainframe security guy's point of view...
1) word is out that the mvs mainframe has never been hacked and no mainframe
security guy wants to be the first to have their system compromised. strong
motivation to keep the troops in line and not buy into any notions of
on JES2
To: IBM-MAIN@bama.ua.edu
Date: Friday, July 29, 2011, 3:52 PM
On Fri, 29 Jul 2011 10:56:58 -0700,
Cris Hernandez #9 hernandez...@yahoo.com
wrote:
but it changes constantly,
got to take the snapshot to see it.
--- On Fri, 7/29/11, Paul Gilmartin wrote:
On Thu, 28 Jul 2011 16
First thing I would do is figure out which statement is causing the error.
I use a simple
lc = RC
to capture the return code from all commands calls (like ALLOC, EXECIO, FREE,
etc.), then code:
IF lc 0 THEN DO
SAY 'Error performing . last RC =' lc
EXIT lc
END
so the exec
well, I use my personal email address and not my work email (I seriously don't
understand how anyone can deal with 50-100 blog emails at work daily), don't
reveal any information related to my OS other than the assumed IBM mainframe
OS, and don't share any code snippets that would provide any
how about making sure the file is created? and put a header record in so it
ain't empty?
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
JCL consists of about 120 words last I checked (not counting those goofy
conditional statements IBM created recently for those who couldn't comprehend
COND codes), average batch job uses about 20 or 25. Seems a whole lot easier
to me to just learn JCL.
contrary to popular misconceptions, JCL is not a programming language, meaning
it's not used to manipulate data. To append data onto any existing file, code
disp=mod instead of shr/new/old and use appropriate pgm/util to read in write
out what you want.
--- On Tue, 7/12/11, Mosley, George
very true from my experience, restarts involve the entire active job tree.
note that batchpipes can be used as both input and output for a single
program/utility, so for maximum speed, the tree can have multiple levels of the
application's batch processing all at once.
also, bottlenecks can
If you have seriously large quantities of data to process in a short amount of
time, it works wonders. used it back when it was new, had 5 jobs reading multi
tape volumes feeding the rest of the 20 batch jobs running various sorts appl
programs. works great when all goes well, a real bear
What command?
Commands are just compiled programs with predefined inputs and outputs.
Output may be written to the terminal, or file allocation named sysprint,
sysout, systsprt, or whatever.
usually what I do is is use REXX code SYSOUTTRAP to capture output written to
terminal then write the
what are you using to open the syslog?
every shop I ever worked the syslog was accessible via whatever JES output
queue viewing tool was in use.
is that what you're talking about?
assuming you're already logging the sysout.
--- On Wed, 6/8/11, Ayhan Yalkut ayhan.yal...@halkbank.com.tr wrote:
please dont think I'm being insultin, but you didnt say...
did you check the obvious?
the physical connections and their corresponding logical definitions (new
library?) ?
--- On Fri, 5/27/11, Ibm Main chm...@gmail.com wrote:
From: Ibm Main chm...@gmail.com
Subject: INOPERATIVE PATH
To:
proper use of mgmtclas/storclas can save a whole lot of dasd and prevents the
need to clean up garbage on a regular basis.
--- On Wed, 5/4/11, ibmnew ibm...@163.com wrote:
From: ibmnew ibm...@163.com
Subject: Do we need to implement HSM
To: IBM-MAIN@bama.ua.edu
Date: Wednesday, May 4,
there is nothing here to loop, me thinks you've mistaken an enqueue/allocation
issue with a loop. ya need a DO WHILE/UNTIL or similar in order to get a loop.
what appears in the systsprnt? and jes logs?
--- On Fri, 4/1/11, Hunkeler Peter (KIUP 4) peter.hunke...@credit-suisse.com
wrote:
incomplete extracts from exec I use to process PDS' but just to supply some
actual rexx code to go with the psycho code:
pds = 'fully.qualified.dsn'
x = OUTTRAP(amems.,,'NOCONCAT')
cmd = LISTD pds MEMBERS
cmd
x = OUTTRAP('OFF')
sellist = ' '
DO i = 7 to amems.0
member =
[mailto:IBM-MAIN@bama.ua.edu]
On Behalf Of Cris Hernandez #9
Sent: Monday, February 14, 2011 4:06 PM
To: IBM-MAIN@bama.ua.edu
Subject: Re: Adding Lines to End of Member
I vote for the REXX option.
Use LISTD after an OUTTRAP to capture a list of
members, DO loop thru the list, allocate the dsn
I vote for the REXX option.
Use LISTD after an OUTTRAP to capture a list of members, DO loop thru the
list, allocate the dsn/member, readin, add the 2 lines to the end by adding to
a counter and setting the needed value to the stemvar, and then write back out.
--- On Mon, 2/14/11,
I'm still reeling from the recent realization that my ISPF skills are stuck in
the dark ages (1980's) but this post went all day without being replied to, so
I'll take this to work and run a few tests before saying anything other than
that the RC value is set to the value of the previous
Written 3270 emulators?
Any emulators out there that not only do the data xfer between mainframe and
pc, but also kick off the pc software required to open the file?
When creating SAS graphs on MVS, I found all the stepx I had to take to view
them on the pc quite annoying.
--- On Tue,
I too have auditors who treat the my mainframe like one those little puters and
I find it best to first educate them before they convince my management to send
me chasing phantoms. Don't assume your auditor won't appreciate a mainframe
education.
The first place to hide a virus is in the OS,
rexx function listdsi ?
--- On Wed, 1/26/11, Roberto Halais roberto.hal...@gmail.com wrote:
From: Roberto Halais roberto.hal...@gmail.com
Subject: Dataset Last Reference Date
To: IBM-MAIN@bama.ua.edu
Date: Wednesday, January 26, 2011, 2:34 PM
Listers:
I have been trying unsuccessfully to
twimc,
and I may be a newbie here,
but I'll vouch for Steve!
Hi Steve,
who knew?
so this is what my sysprogs are doing all day long that keeps them too busy to
return my emails...
hernandez
--- On Fri, 1/21/11, Steve Conway steve_con...@ao.uscourts.gov wrote:
From: Steve Conway
COND=EVEN and COND=ONLY only take effect when there is a system or user abend
returned, they have nothing to do with cond code checking.
If either failed to perform according to specs, I'd be raising cain with my
sysprogs in short order.
hernandez
--- On Fri, 1/21/11, Stan Weyman
suppose I'm old school, never saw a need to code IF-THEN-ELSE in JCL, only use
cond code checking and never have any issues with step execution or job flow.
the only cond code testing I ever do when writing procs is, if it's true, it's
through, meaning the step/job won't execute if the COND is
I haven't seen this happen since we disposed of our last punch card reader!
but even then, all that needed to be done was send a delimiter ( // only )
thru the machine and hatever was stuck was immediately submitted in whatever
shape it was in JCL-wise. With today's connections, don't know of
huh, never had that issue...
maybe a spf macro (not my area of expertise) using edit command:
f p'.'
or
c p'.' ' ' all
is there only 1 character or multiple characters that your print function can't
handle?
don't have the rexx manual handy, but there are hex conversion functions the
could
don't know what you're trying to do with it, but for transport purposes, I run
a utility on the mainframe to make a flatfile then xmit as text. no corruption
when sent back to the mainframe if receiving file is pre-allocated with correct
LRECL RECFM. IEBUPDTE (or is it IEBCOPY?) or SAS Proc
another reason why raw instream data is a bad idea... if more than one person
edits the JCL, one has spf profile with 'num on' and another has 'num off', you
can seriously mess up your data when editing in spf.
--
: Re: passing instream sysin to next step
To: IBM-MAIN@bama.ua.edu
Date: Wednesday, January 12, 2011, 12:56 PM
On Tue, 11 Jan 2011 15:44:20 -0800,
Cris Hernandez #9 wrote:
I agree with this post. Why read the same file twice
other than to make a copy of it, especially when it's
something
On Wed, Jan 12, 2011 at 3:01 PM, Cris
Hernandez #9
hernandez...@yahoo.comwrote:
ok, let's back up to the beginning. why would
anyone want to put their
data into a SYSIN DD * ?
doesn't sound like anything that would happen in a
production environment,
or QA testing, so that leaves
I agree with this post. Why read the same file twice other than to make a copy
of it, especially when it's something as easily corruptible as raw, instream
data coded within JCL.
hernandez
--- On Tue, 1/11/11, Lizette Koehler stars...@mindspring.com wrote:
From: Lizette Koehler
some things aren't meant to be done foreground. why not run a batch job to
sort the file into 2 files, one containing records requiring edit, and the
other the ones that don't require editing. make the changes and merge back
together batch mode. if all records need editing, write a batch
60 matches
Mail list logo