There are no coding requirements for the application, When you do
a QSAM OPEN for Input, the first read-ahead I/Os are scheduled by OPEN,
and the application program can proceed without waiting after the OPEN at
least to the point of doing the first GET. Subsequent read-ahead I/Os can
ov
I'm bowled over by David Noon's post. I did not know that QSAM allowed
asynchronous I/O operations and have not looked into coding requirements.
At the same time I contend that system managed interleaving is not the same
thing. While it undoubtedly speeds up I/O for 'traditional' QSAM, the
app
QSAM does I/O interleaving. The exact amount of buffers required to
trigger the behavior has varied somewhat over the years. 20 - 30 seems to
be where it settled. Haven't looked into it in years.
Rob Schramm
On Fri, Feb 3, 2017, 8:15 PM Edward Gould wrote:
> > On Feb 3, 2017, at 6:27 PM, Dav
> On Feb 3, 2017, at 6:27 PM, David W Noon
> <013a910fd252-dmarc-requ...@listserv.ua.edu> wrote:
>
> On Fri, 3 Feb 2017 16:54:57 -0600, Paul Gilmartin
> (000433f07816-dmarc-requ...@listserv.ua.edu) wrote about "Re: BSAM
> vs QSAM" (in <4092052368963851.wa.paulgboulderaim@listserv.ua.e
We do lots of alter for mgmtclas. We cannot alter Dataclass. Since this
question relates to mgt, I am fairly certain it will work.
Other option is batch migration command
HMIG dsname ML2 NOWAIT
Lizette
> -Original Message-
> From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTS
On Sat, 4 Feb 2017 00:27:38 +, David W Noon wrote:
>
>> Another concern if you need to support BPAM is that BPAM and BSAM can share
>> more
>> code than BPAM and QSAM.
>
>That's fairly marginal. Much of SAM/E is in the LPA.
>
I was thinking of source code, from the viewpoint of a developer of
Liz:
This is going back many years but when GUIDE last asked IBM to do this they
said it wasn’t possible.
Has IBM reversed their position?
Ed
> On Feb 3, 2017, at 5:56 PM, Lizette Koehler wrote:
>
> You can do an ALTER name MGMTCLAS(x).
>
> Once the MGMTClass of your dreams is in SMS/HSM, you
On Fri, 3 Feb 2017 16:54:57 -0600, Paul Gilmartin
(000433f07816-dmarc-requ...@listserv.ua.edu) wrote about "Re: BSAM
vs QSAM" (in <4092052368963851.wa.paulgboulderaim@listserv.ua.edu>):
> On Fri, 3 Feb 2017 14:42:24 -0500, Farley, Peter x23353 wrote:
>>
>> OTOH you don't have to wait for c
You can do an ALTER name MGMTCLAS(x).
Once the MGMTClass of your dreams is in SMS/HSM, you can then do an ALTER
command (use the ISMF Panels or create a batch IDCAMS) to alter the datasets to
the new class.
The ALTER is effective as soon as it is done.
Then you will need to verify in the Stora
Not sure if QSAM overlaps any I/O or not. You would think so, but TFM has
little to say about it that I found so far.
Peter
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf
Of Paul Gilmartin
Sent: Friday, February 03, 2017 5:55 PM
To: I
For a specific file prefix, my STORCLAS ACS routine sets the storage
class to SC. Then in my STORGRP ACS routine, that storage class
causes the storage group to be set to SG.
But, in my MGMTCLAS ACS routine, where everything is based on the
storage group setting, there are no specific
From the application program's point of view, there is no overlap. Once a read
or write is issued, the TCB WAITs until the I/O is complete. Of course the
application can code for subtask processing to do I/O separately from
calculation, but that's a whole nother level of escalation in program
c
On Fri, 3 Feb 2017 14:42:24 -0500, Farley, Peter x23353 wrote:
>
>OTOH you don't have to wait for completion of a READ or a WRITE. You can
>issue a WRITE at the end of a processing loop and then go back to process the
>next record while the WRITE completes, and only CHECK the WRITE when you are
You're welcome.
Beware that I used "record" in my note to mean "block". The assumption being
that you process all the records in a READ block or construct all the records
in a WRITE block before issuing the next READ or WRITE, respectively.
I remember one notably complex program I was exposed
Thanks so much for the tip
> On Feb 3, 2017, at 2:42 PM, Farley, Peter x23353
> wrote:
>
> BSAM only gets you an entire block on a READ. You have to extract each
> varying record from the block with your own code.
>
> On a WRITE you have to give it an entire block, BDW + one or more RDW +
When reading the doc, be careful about the term 'record'. For a physical
device, 'record' means 'block' because the device has no notion of logical
structure. As already pointed out, BSAM read returns an entire block (as in
BLKSIZE) and leaves the task of segmenting into logical records entirely
I was assuming distinct read and write buffers. I have used LOCATE mode on
READ's and WRITE's to avoid just that issue. That's another potential speed-up
using BSAM, using LOCATE mode means you don't have to move large blocks of data
around (given how slow MVCL(E) operations are.
Obviously it
Careful on reusing that buffer before the ECB is posted in the write. You
could wind up overlaying what is in the middle of being written. It is not
safe to reuse the buffer before the ECB is posted. You may get away with it
most of the time, but if there is an error and the channel re-drives
I was working an issue with IBM support via a PMR and during the PMR I posted a
URL to a knowledge base document and they posted one for me.
Both URL's wrapped and were not usable without some guesswork and editing as
the hyperlink was broken in the wrap.
I reported this in the PMR and was told
If you issue a BSAM READ followed by a WAIT, and then deblock the buffer before
doing the next READ, the system builds a channel program to read just the one
block.
If on the other hand you issue 10 READ commands in a loop, each READ having its
own DECB, the system will build a channel program t
BSAM only gets you an entire block on a READ. You have to extract each varying
record from the block with your own code.
On a WRITE you have to give it an entire block, BDW + one or more RDW + record.
You have to construct the block yourself in your own code before you issue the
WRITE.
OTOH
> -Original Message-
> From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On
> Behalf Of Joseph Reichman
> Sent: Friday, February 03, 2017 11:27 AM
> To: IBM-MAIN@LISTSERV.UA.EDU
> Subject: Re: BSAM vs QSAM
>
> I have huge VB files
>
> Don't really understand what you m
I have huge VB files
Don't really understand what you mean by
Deblock after doing a READ then WAIT
Where an entire block is read subsequent READs
Just point to the next record
> On Feb 3, 2017, at 2:22 PM, Blaicher, Christopher Y.
> wrote:
>
> There can be if you code for just what you ex
I would say that BSAM is a lot more complicated than QSAM. Whether the small
performance advantage is worth extra debugging effort and disruption to
production has to be factored in to the choice. In my personal view, it's
almost never worth the extra hassle.
Note however that some very specifi
There can be if you code for just what you expect.
QSAM does multi-buffer I/O for you, with BSAM you have to issue multiple WRITE
or REAAD commands and do a WAIT, not to mention having to block or de-block the
buffer, which can be a real pain for VBS files.
It really depends on how much you are
Hi
BSAM is a bit more complex than QSAM
Is there any performance improvement
Thanks
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
>I think it would be useful to document this absence of practical
limitations.
What you are really talking about is not the absence of a practical
limitation but the absence of any enforced or expected limit. There is
(always) a practical limit. That limit is whatever would cause "too much"
re
On some products (e.g. Mainview) we license x MIPS and send SCRT reports to BMC
monthly. As long as our peak 4HA for the year is under our license we are OK.
Usage for Control/M is based on reports from Control/M on number of jobs run.
CIT | Ken Porowski | VP Mainframe Engineering | Informat
You mentioned Usage. Interestingly enough, at least to me, is that BMC is
one of the vendors that doesn't seem to use IFAUSAGE - so no Usage Data
sections in SMF 30 and (I would guess) nothing in SMF 89.
I'm not sure what to make of that, except to assume their licencing works
differently.
Any
Thanks for the info Ken. That's one of our worries with the increased MIPS our
serial will still be the same but the model number will obviously change and we
are not on any capacity or usage charge so we have no idea how the products
will react with the change in model and if they will still co
BMC does have sub-capacity licensing so I highly doubt they check total MIPS
but I believe they are model and serial dependent.
Our Control/M license is usage based (number of jobs) not MIPS based.
CIT | Ken Porowski | VP Mainframe Engineering | Information Technology | +1 973
740 5459 (tel) |
Hi,
I'm looking to find out if anyone knows if BMC products make use of CPU Serial
alone or MIPS and CPU serial for it's licensing checks?
We're planning on adding MIPS to the system by moving to a higher config for an
additional LPAR but the added LPAR will have no BMC products and we current
Yes, that would work, but then I have a program maintenance issue should
anything in the generated JCL change.
Couple that with the fact that the catalog extract portion which is generating
the INCLUDE member is targeted to be used by more than a single job, I would
have to multiple copies of th
Chuck,
Why use INCLUDE MEMBER= in a second job? A sequential dataset will do just as
well to build the entire JCL you need, including the dynamically generated tape
allocations, then just write the whole thing to an internal reader. KISS
principle.
Peter
-Original Message-
From: IBM
Standard SSH/SFTP doesn't support X.509 certificate's for authentication,
so I don't understand your reference to a CA.
(z/OS OpenSSH does allow you to put SSH public and private keys in a Key
Ring Certificate, but only the keys are used; the certificate and its
signature are irrelevant.)
Kirk Wo
On Wed, 1 Feb 2017 07:51:23 -0600, Kirk Wolf wrote:
>> Remember that although the integrity of public keys needs to be guarded,
>their privacy does not.
>So it is common to use other secure communications, like publishing the
>public key on a https: web page.
The issue I have with that is one of
Ahhh!
Thanks John!
I hadn't gotten to the DYNALLOC support yet. That was the agenda for this
morning.
It appears that I can't "get there from here" as the saying goes.
I have thought of another means of doing what I what, I just didn't want to do
it.
I was hoping to have my solution be a single
37 matches
Mail list logo