Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

An OSA TN3270 question

416 views
Skip to first unread message

Tobias, Sid

unread,
Mar 14, 2010, 8:37:08 AM3/14/10
to

Currently we use a Cisco TN3270 CIP card in a router which has an ESCON connector.
It has been working great for 6 years and we will continue to use it at our main site.
We are currently building a D/R site and would like to have an alternative to the CISCO solution.
Mainly because CISCO does not sell them or support them anymore. I have read that the reason that CISCO doesnt make them anymore is that IBM offers the OSA that is as reliable and as fast as the CIP Card solution.

My question is, has anyone used the OSA to support 2000 TN3270 sessions?  If so how does it interface to CICS (VTAM or TCP).  If anyone has done this i would appreciate seeing the OSA parameters and or the VTAM parameters for this.


Thank you


Sid Tobias
Health Management Associates, Inc.
Computer Operations Manager
239 598 4669 x 3462




Rich Smrcina

unread,
Mar 14, 2010, 8:43:47 AM3/14/10
to
I don't know about supporting 2000 sessions and I don't have the OSA
definitions handy. To the operating system (VM, VSE or z/OS) the
terminal session looks like a non-SNA terminal session.

There is a good reference on the OSA ICC on the redbooks site:

OSA-Express Integrated Console Controller Implementation Guide
SG24-6364-01

--
Rich Smrcina
Phone: 414-491-6001
http://www.linkedin.com/in/richsmrcina

Catch the WAVV! http://www.wavv.org
WAVV 2010 - Apr 9-13, 2010 Covington, KY

Wakser, David

unread,
Mar 14, 2010, 8:58:56 AM3/14/10
to

Sid:

 

                You could set it up either way; by using a different port (and by specifying the application desired in the DEFINE TELNETD statements for that port), you could send TCP/IP users directly into CICS, or any other VTAM application. You could also set up a TCP/IP menu from which they can choose their sessions. To the system, the in-coming connection is plain old non-SNA. Except for the TCP/IP definition for the device, there is no difference between a CISCO or OSA connection to the system.

 

                For the large number of connections, you need to ensure that you have supplied plenty of partition GETVIS for your TELNET connections in the TCP/IP partition. On one of our systems, we have defined that partition as 50M – but you may need to adjust that. Otherwise, there should be no surprises.

 

David Wakser

Confidentiality Note: This e-mail, including any attachment to it, may contain material that is confidential, proprietary, privileged and/or "Protected Health Information," within the meaning of the regulations under the Health Insurance Portability & Accountability Act as amended. If it is not clear that you are the intended recipient, you are hereby notified that you have received this transmittal in error, and any review, dissemination, distribution or copying of this e-mail, including any attachment to it, is strictly prohibited. If you have received this e-mail in error, please immediately return it to the sender and delete it from your system. Thank you.

Rbo...@aol.com

unread,
Mar 14, 2010, 9:08:31 AM3/14/10
to
Sid,
 
OSA-ICC only supports about 255 sessions. I forget the exact number but it's no where near 2000. In my case I setup 27 sessions (3 sessions for 9 LPAR's). One of the 3 sessions, per LPAR, is a CONSOLE and the other 2 are VTAM LU's, per LPAR.  The VTAM LU's look exactly like normal VTAM LU's. Here's an example of one:
 
L1GO401O LOCAL CUADDR=401,                                             X       
               TERM=3277,                                              X       
               FEATUR2=(EDATS,MODEL2),                                 X       
               MODETAB=ISTINCLM,                                       X       
               USSTAB=VTMUSST3,                                        X       
               DLOGMOD=D4B32XX3
 
The IOCP definitions are as follows:
 
         CHPID PCHID=121,PATH=(73),TYPE=OSC,SHARED OSA-ICC
         CNTLUNIT CUNUMBR=731,PATH=(73),UNIT=OSC OSA-ICC          
         IODEVICE ADDRESS=(400,32),CUNUMBR=(731),UNIT=3270,MODEL=X
  
Connectivity over and above the above is done through TCP/IP for ALL other sessions.
 
The remainder of the setup is done using the HMC. It's pretty simple to setup. You define a server and then the sessions, where you specify that each sessions is to be a CONSOLE or TERMINAL. Here's an example of the output that's produced when downloading the each sessions parameters (it illustrates the server, one console and one terminal):
 
// This file has been generated from the binary file D:\POKCODE\IQZC0121.HUL
 
<OSC_SERVER>
  HOST_IP= 10.254.38.21
  DEFAULT_GATEWAY= 10.254.38.1
  SUBNET_MASK= 255.255.255.0
  PORT= 3270
  ETHERNET_FRAME= DIX
  MTU= 1492
  NAME= CDCz890
</OSC_SERVER>
 
 
<CONFIG_SESSION>
<SESSION1>
  CSS=  00 IID=  01 DEVICE=  0400
  GROUP=  "KEV1OP"
  CONSOLE_TYPE=  2    RESPONSE=  ON
  READ_TIMEOUT=  10
  DEFER_HOST_DISCONNECT=  0
</SESSION1>
 
<SESSION2>
  CSS=  00 IID=  01 DEVICE=  0401
  GROUP=  "KEV1LO1"
  CONSOLE_TYPE=  1    RESPONSE=  OFF
  READ_TIMEOUT=  60
  DEFER_HOST_DISCONNECT=  60
</SESSION2>
 
 
In a message dated 3/14/2010 8:43:37 A.M. Eastern Daylight Time, ri...@velocitysoftware.com writes:
On 03/14/2010 07:34 AM, Tobias, Sid wrote:
>
> Currently we use a Cisco TN3270 CIP card in a router which has an
> ESCON connector.
> It has been working great for 6 years and we will continue to use it
> at our main site.
> We are currently building a D/R site and would like to have an
> alternative to the CISCO solution.
> Mainly because CISCO does not sell them or support them anymore. I
> have read that the reason that CISCO doesnt make them anymore is that
> IBM offers the OSA that is as reliable and as fast as the CIP Card
> solution.
>
> My question is, has anyone used the OSA to support 2000 TN3270
> sessions?  If so how does it interface to CICS (VTAM or TCP).  If
> anyone has done this i would appreciate seeing the OSA parameters and
> or the VTAM parameters for this.
>
>
> Thank you
>
>
> Sid Tobias
> Health Management Associates, Inc.
> Computer Operations Manager
> 239 598 4669 x 3462
>
>
>
>

Kevin Corkery

unread,
Mar 14, 2010, 9:16:40 AM3/14/10
to
The OSA-ICC supports a maximum of 120 TN3270 session per port.  They can be pooled on the port but not across ports.  This is the local 3270 console support only.  I have 2 ports, on different cards, designated for OSA-ICC with 240 terminal sessions available.  For a small shop this is more than enough.

From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Rbo...@aol.com
Sent: Sunday, March 14, 2010 9:08 AM
To: VSE Discussion List
Subject: Re: An OSA TN3270 question

Wakser, David

unread,
Mar 14, 2010, 9:20:58 AM3/14/10
to

Bob:

 

                Perhaps I misunderstood, but I didn’t think he was talking about ICC using the HMC. I assumed he was merely speaking about an OSA card connected to the VSE system.

 

David Wakser

 

From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Rbo...@aol.com
Sent: Sunday, March 14, 2010 9:08 AM
To: VSE Discussion List
Subject: Re: An OSA TN3270 question

 

Sid,

 

OSA-ICC only supports about 255 sessions. I forget the exact number but it's no where near 2000. In my case I setup 27 sessions (3 sessions for 9 LPAR's). One of the 3 sessions, per LPAR, is a CONSOLE and the other 2 are VTAM LU's, per LPAR.  The VTAM LU's look exactly like normal VTAM LU's. Here's an example of one:

 

Confidentiality Note: This e-mail, including any attachment to it, may contain material that is confidential, proprietary, privileged and/or "Protected Health Information," within the meaning of the regulations under the Health Insurance Portability & Accountability Act as amended. If it is not clear that you are the intended recipient, you are hereby notified that you have received this transmittal in error, and any review, dissemination, distribution or copying of this e-mail, including any attachment to it, is strictly prohibited. If you have received this e-mail in error, please immediately return it to the sender and delete it from your system. Thank you.

Rbo...@aol.com

unread,
Mar 14, 2010, 9:23:04 AM3/14/10
to
David,
 
Maybe "I" was the one that misunderstood. Oh well, it looks like he got answers for both.

Chris Mason

unread,
Mar 14, 2010, 1:04:27 PM3/14/10
to
Sid
 
> Mainly because CISCO does not sell them or support them anymore, I have read that the reason that CISCO doesn't
 
make them anymore is that IBM offers the OSA that is as reliable and as fast as the CIP Card solution.
 
Perhaps you can let us know *where* you read that. It's "apples and oranges".
 
I believe that the CIP "solution" is to provide a general-purpose TN3270 function - perhaps among other capabilities. That would mean comparing the "CIP" with "OSA" was not comparing "like with like".
 
The flavour of "OSA" generally abbreviated to the OSA-ICC[1] running as CHPID type OSC is oriented to providing support for a hardware console which, because it presents the appearance of a local pre/non-SNA channel-attached 3270 display, can also be used incidentally as a 3270 display in an SNA network by means of "protocol conversion" logic in the VTAM running in the LPAR to which the "channel" is defined. This is *not* a general-purpose TN3270 server as the limits on the number of TN3270E TCP connections - illogically and inappropriately sometimes described as "sessions" - mentioned by Bob Botsis and Kevin Corkery indicates.
 
The more general purpose flavours of OSA, typically running in QDIO mode as CHPID type OSD, behave as LAN, typically Ethernet, adapters and can support TN3270 or TN3270E connections only when *additionally* supported by a TN3270(E) server function running as an application to the IP-support software on the operating system, in the case of this list, VSE.[2]
 
> My question is, has anyone used the OSA to support 2000 TN3270 sessions?
 
I expect that using an OSA as a LAN adapter where the TN3270(E) server support runs as an operating system application can extraordinarily easily support 2,000 TCP connections and concatenated SNA sessions.
 
> If so how does it interface to CICS (VTAM or TCP).
 
If you pass through a TN3270 server - as you do with the CIP - then the TN3270 server creates the appearance of a 3270 display. It can either pretend to be a pre/non-SNA channel-attached 3270 display or it can pretend to be the secondary half-session in an SNA LU-LU session using 3270 data streams following either LU type 0 or LU type 2 protocols.
 
If you expect not to change what the client workstations do and how they are defined, then you are stuck with TN3270 or TN3270E clients and a TN3270(E) server will be required as the partner in the TN3270(E) TCP connection.
 
If you pass through a TN3270(E) server, the interface to CICS will almost certainly be over the VTAM API using CICS's VTAM/SNA support.
 
Total and complete honesty prompts me to say that, if you use a TN3270(E) server which supports pre/non-SNA channel-attached 3270 display devices, at one time and just possibly still today, CICS could support you more directly though BTAM - although I'm not sure of the position of BTAM support in the various z operating systems these days.
 
That *may* be the answer to the "interface to CICS with TCP". If not, please post again.
 
-
 
Quoting David Wakser
 
> ... there is no difference between a CISCO or OSA connection to the system.
 
I checked on a web site for "TN3270 CISCO CIP"[3] and found that the "dynamic definition of dependent LUs", DDDLUS, function was supported. This implies that the CIP emulates an SSCP-dependent PU entity and associated SSCP-dependent LU entities - the dynamic appearance of which is what the DDDLUS function is all about - which are defined within VTAM using a PU statement and, if not relying entirely on the DDDLUS function, following LU statements.[4]
 
If using an OSA-ICC, the LU entity definitions within VTAM are LOCAL statements. If using an OSA as a LAN adapter together with a TN3270(E) server application, the LU entity definitions within VTAM are APPL statements - or VSE has clever ways of dealing with these matters which are foreign to me!
 
So, unless David can flesh out what he means by "no difference" you shouldn't expect to be using the same VTAM definitions with one of the two "OSA" "solutions" as with your "CIP" "solution".
 
-
 
Quoting Bob Botsis (RBOTSIS)
 
> The VTAM LU's look exactly like normal VTAM LU's.
 
***Only*** if your "normal" VTAM LU is supported by the "protocol conversion" logic within VTAM for pre/non-SNA channel-attached 3270 devices which are represented to VTAM sing LOCAL statements rather than real LU statements or perhaps APPL statements.
 
Anyhow we can that Bob for his samples.
 
I *think* the code identified as CONSOLE_TYPE refers to the last time I got extensively involved in a discussion in this list where I suggested scepticism over what the good people responsible for the OSA-ICC said about how this code should be defined. It may be useful to read over that thread[5] - if you decide to use the OSA-ICC - which appears unlikely.
 
-
 
Chris Mason
 
[1] OSA-Express Integrated Console Controller - as indicated by Rich Smrcina.
 
[2] I'd better say I haven't been a VSE specialist since 1969 - when it was known as DOS! I'm here to pick up VTAM and SNA topics - and maybe IP topics as long as they don't involve the actual IP-based software running with VSE.
 
 
[4] If relying wholly or partially on the DDDLUS function, there will be some *model* LU statements defined.
 
----- Original Message -----
Sent: Sunday, March 14, 2010 1:34 PM
Subject: An OSA TN3270 question

Wakser, David

unread,
Mar 14, 2010, 1:15:06 PM3/14/10
to

Chris and Sid:

 

                We had, in the past, one client where we used a CIP device, because we were having problems with their SNA connections on an OSA-2, and IBM was unable to diagnose the problem. Therefore, we migrated them to a CIP, which allowed both traditional SNA connections along with TCP/IP connections. However, I was not involved with the hardware or VTAM side of the implementation, and the client has since migrated to a different platform, so I cannot access the VTAM settings used. However, from a TCP/IP and TN3270 side (which I set up and supported), there was no difference between the CIP and standard (not OSA-ICC) OSA-2.

 

David Wakser

 

From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Chris Mason
Sent: Sunday, March 14, 2010 1:04 PM
To: VSE Discussion List
Subject: Re: An OSA TN3270 question

 

Sid

Confidentiality Note: This e-mail, including any attachment to it, may contain material that is confidential, proprietary, privileged and/or "Protected Health Information," within the meaning of the regulations under the Health Insurance Portability & Accountability Act as amended. If it is not clear that you are the intended recipient, you are hereby notified that you have received this transmittal in error, and any review, dissemination, distribution or copying of this e-mail, including any attachment to it, is strictly prohibited. If you have received this e-mail in error, please immediately return it to the sender and delete it from your system. Thank you.

Edward M Martin

unread,
Mar 14, 2010, 7:45:23 PM3/14/10
to

Hello Sid,

 

Question, do you have z/VM or are straight z/VSE?

 

We have z/VM and TCP/IP with a  limit of 4096 on the z/VM side.

On the z/VSE direct connections we have about 1500 terminals and 900 printers with a 70 meg partition. We have a separate TCP/IP partition for all the email/ftp stuff.

 

 

Ed Martin

Aultman Health Foundation

330-363-5050

ext 35050

From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Tobias, Sid


Sent: Sunday, March 14, 2010 8:35 AM
To: VSE Discussion List

Chris Mason

unread,
Mar 15, 2010, 6:05:24 AM3/15/10
to
David
 
This makes sense only by assuming your memory has been playing trick with you!
 
You are still trying to equate OSA features, even OSA-2 (not OSA-ICC), with CIP supporting TN3270. OSA features, OSA-ICC excepted, know nothing about TN3270. If the OSA appears to be supporting TN3270, either TCP traffic over IP flows back and forth through to an "inboard" TN3270(E) server application - or - SNA traffic flows back and forth through to an "outboard" implementation of TN3270(E) logic in some clever typically router-like machine.
 
If you take yourself to a "bookshelf" where the OSA feature manuals are to be found such as
 
 
and search for the character string "TN3270E", you will find a "hit" only in the manual for the OSA-ICC, "System z9 and eServer zSeries z890 and z990 Open Systems Adapter-Express Integrated Console Controller User's Guide". You will find a hit in neither "System z10, System z9 and eServer zSeries Open Systems Adapter-Express Customer's Guide and Reference" nor "eserver zSeries 900 Planning for the Open Systems Adapter-2 Feature", the two manuals which cover the OSA not used as an OSA-ICC and OSA-2 respectively.
 
I suspect - on somewhat flimsy evidence - that the VTAM definitions which you found appeared to be being used with both the OSA-2 and CIP related to, in the case of the OSA-2, the SNA connections to which you referred to workstations themselves using SNA support - and, in the case of CIP, possibly the same or CIP supporting TN3270. In all these cases, there would be a PU statement followed, unless the DDDLUS function were being used exclusively, by LU statements.
 
In the case of the "inboard" TN3270(E) application I cannot imagine it could be done in any way except one very similar to the way it is done with the z/OS Communications Server IP component which involves VTAM APPL statements representing the secondary LU in the session using the 3270 data stream.
 
In the case of the "outboard" TN3270(E) implementation, VTAM LU statements represent the secondary LU in the session using the 3270 data stream.
 
Just for completeness, in the case of "hardware" which simulates the "upstream" appearance of a pre/non-SNA channel-attached 3270 controller - originally the 3272 - such as the OSA-ICC, VTAM LOCAL statements represent the secondary LU in the session using the 3270 data stream - and imply the use of the VTAM protocol converter for non-SNA 3270 data stream flow.
 
Here are some web pages which helped me clarify what Sid's current VTAM definitions probably are:
 
TN3270 Server for CIP and CPA
 
 
TN3270 Cisco SNA Direct PU Configuration and Verification
 
 
Configuring and Verifying TN3270 CSNA DLUR/DLUS
 
 
Chris Mason

Tobias, Sid

unread,
Mar 15, 2010, 9:47:50 AM3/15/10
to

Chris

No sense in guessing, these are the Vtam Definitions used for the CIP.

 

 

CISCOXCA   VBUILD TYPE=XCA                                             

CISXCA   PORT   ADAPNO=0,CUADDR=700,SAPADDR=04,MEDIUM=RING,TIMER=60    

CISGRP   GROUP ANSWER=ON,                                              X

               AUTOGEN=(100,L,P),                                      X

               CALL=INOUT,                                             X

               DIAL=YES,                                               X

               ISTATUS=ACTIVE                                          

/+                                                                     

CATALOG CISCOSWN.B   R=YES                                              

CISCOSWN   VBUILD TYPE=SWNET,MAXGRP=10,MAXNO=256,MAXDLUR=256           

XCAPU1   PU    ADDR=01,VPACING=0,                                      X

               IDBLK=05D,IDNUM=00700,                                  X

               MAXPATH=256,DISCNT=NO,MAXOUT=7,                         X

               LUGROUP=XCALU1,LUSEED=XCAL1###,                         X

               PUTYPE=2,MAXDATA=256                                    

XCAPU2   PU    ADDR=01,VPACING=0,                                      X

               IDBLK=05D,IDNUM=00701,                                  X

               MAXPATH=256,DISCNT=NO,MAXOUT=7,                         X

               LUGROUP=XCALU1,LUSEED=XCAL2###,                         X

               PUTYPE=2,MAXDATA=256                                    

XCAPU3   PU    ADDR=01,VPACING=0,                                      X

               IDBLK=05D,IDNUM=00702,                                  X

               MAXPATH=256,DISCNT=NO,MAXOUT=7,                         X

               LUGROUP=XCALU1,LUSEED=XCAL3###,                         X

               PUTYPE=2,MAXDATA=256                                     

XCAPU4   PU    ADDR=01,VPACING=0,                                      X

               IDBLK=05D,IDNUM=00703,                                  X

               MAXPATH=256,DISCNT=NO,MAXOUT=7,                         X

               LUGROUP=XCALU1,LUSEED=XCAL4###,                         X

               PUTYPE=2,MAXDATA=256                                     

XCAPU5   PU    ADDR=01,VPACING=0,                                      X

               IDBLK=05D,IDNUM=00704,                                  X

               MAXPATH=256,DISCNT=NO,MAXOUT=7,                         X

               LUGROUP=XCALU1,LUSEED=XCAL5###,                         X

               PUTYPE=2,MAXDATA=256                                     

XCAPU6   PU    ADDR=01,VPACING=0,                                      X

               IDBLK=05D,IDNUM=00705,                                  X

               MAXPATH=256,DISCNT=NO,MAXOUT=7,                         X

               LUGROUP=XCALU1,LUSEED=XCAL6###,                         X

               PUTYPE=2,MAXDATA=256                                      

/+                                                                      

 

CISCOLUS VBUILD TYPE=LUGROUP                                           

*                                                                      

XCALU1   LUGROUP                                                        

*                                                                      

327802   LU    DLOGMOD=D4C32782,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327803   LU    DLOGMOD=D4C32783,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327804   LU    DLOGMOD=D4C32784,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327805   LU    DLOGMOD=D4C32785,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

327902   LU    DLOGMOD=D4C32782,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327903   LU    DLOGMOD=D4C32783,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327904   LU    DLOGMOD=D4C32784,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327905   LU    DLOGMOD=D4C32785,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327802E  LU    DLOGMOD=SNX32702,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327803E  LU    DLOGMOD=SNX32703,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327804E  LU    DLOGMOD=SNX32704,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

327805E  LU    DLOGMOD=SNX32705,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327902E  LU    DLOGMOD=SNX32702,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327903E  LU    DLOGMOD=SNX32703,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327904E  LU    DLOGMOD=SNX32704,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

327905E  LU    DLOGMOD=SNX32705,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3278S2   LU    DLOGMOD=D4C32782,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3278S3   LU    DLOGMOD=D4C32783,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

3278S4   LU    DLOGMOD=D4C32784,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3278S5   LU    DLOGMOD=D4C32785,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3279S2   LU    DLOGMOD=D4C32782,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3279S3   LU    DLOGMOD=D4C32783,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3279S4   LU    DLOGMOD=D4C32784,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3279S5   LU    DLOGMOD=D4C32785,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3278S2E  LU    DLOGMOD=SNX32702,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

3278S3E  LU    DLOGMOD=SNX32703,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                           

3278S4E  LU    DLOGMOD=SNX32704,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

3278S5E  LU    DLOGMOD=SNX32705,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

3279S2E  LU    DLOGMOD=SNX32702,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

3279S3E  LU    DLOGMOD=SNX32703,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

3279S4E  LU    DLOGMOD=SNX32704,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                             

3279S5E  LU    DLOGMOD=SNX32705,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

@        LU    DLOGMOD=D4C32782,                                       X

               MODETAB=ISTINCLM,                                       X

               USSTAB=VTMUSSTR,                                        X

               SSCPFM=USSSCS                                            

/+                                                                      

 

 

 

 

Sid Tobias

Health Management Associates, Inc.

Manager, Computer Operations

239 552 3462

Wakser, David

unread,
Mar 15, 2010, 10:02:42 AM3/15/10
to

No argument from me – as I mentioned, I have no access to the VTAM books that were being used at the time.

 

David Wakser

 

From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Chris Mason


Sent: Monday, March 15, 2010 6:05 AM

To: VSE Discussion List
Subject: Re: An OSA TN3270 question

 

David

 

This makes sense only by assuming your memory has been playing trick with you!

 

Confidentiality Note: This e-mail, including any attachment to it, may contain material that is confidential, proprietary, privileged and/or "Protected Health Information," within the meaning of the regulations under the Health Insurance Portability & Accountability Act as amended. If it is not clear that you are the intended recipient, you are hereby notified that you have received this transmittal in error, and any review, dissemination, distribution or copying of this e-mail, including any attachment to it, is strictly prohibited. If you have received this e-mail in error, please immediately return it to the sender and delete it from your system. Thank you.

Tobias, Sid

unread,
Mar 15, 2010, 10:23:23 AM3/15/10
to

Ed

We have Z/VM 5.4 and Z/VSE 4.2 and TCP/IP.

We have a Z/10 P01 in production   and the D/R machine will be a  Z/10 E-10 with CBU.

On your terminal side what percent of the CPU is TCP using.

 

Sid Tobias

Health Management Associates, Inc.

Manager, Computer Operations

239 552 3462

 

From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Edward M Martin


Sent: Sunday, March 14, 2010 7:45 PM
To: VSE Discussion List

Subject: RE: An OSA TN3270 question

 

Hello Sid,

indust...@winwholesale.com

unread,
Mar 15, 2010, 11:02:19 AM3/15/10
to
owner...@Lehigh.EDU wrote on 03/14/2010 08:34:49 AM:
> My question is, has anyone used the OSA to support 2000 TN3270
> sessions?  If so how does it interface to CICS (VTAM or TCP).


        If you want the OSA to supply TN3270 support, then that would be OSA-ICC (OSC in the IOCDS) and would have the session limits already described.  On the VSE side, these would be defined as individual 3277 IPL addresses and to VTAM as local non-SNA terminals.  No VSE TCP/IP stack is used.

        If you just want the OSA to supply the LAN interface, then that would generally be OSA-2 (OSE in the IOCDS) or OSA-X (OSD in the IOCDS) and would support more like 4096 sessions per OSA port.  On the VSE side, these would be defined as dual  OSA IPL addresses or a triplet of OSAX IPL addresses and to both your VSE TCP/IP stack and to VTAM or BSI's ITAM.

        Definitions supplied on demand.   ;-)

Sincerely,

Dave Clark

WinWholesale Group Services
3110 Kettering Boulevard
Dayton, Ohio  45439  USA
(937) 294-5331


This email message and any attachments is for use only by the named addressee(s) and may contain confidential, privileged and/or proprietary information. If you have received this message in error, please immediately notify the sender and delete and destroy the message and all copies. All unauthorized direct or indirect use or disclosure of this message is strictly prohibited. No right to confidentiality or privilege is waived or lost by any error in transmission.

Chris Mason

unread,
Mar 15, 2010, 12:55:51 PM3/15/10
to
Sid
 
Thanks for the confirmation of the CIP configuration.
 
CIP simulates a 3172 connected to a token ring LAN internal to the CIP which then supports TN3270E TCP connections to a TN3270E server as if they were grouped into sets of SSCP-dependent LUs under an SSCP-dependent PU on one or more, 6 actually configured in your case, SNA nodes.
 
Moreover I see you have made extensive use of the DDDLUS function where individual LU definition statements, each with a LOCADDR operand, following the PU statement are replaced by LU model statements. It appears that the character string required for the DDDLUS function is taken from the device type codes specified in RFC 2355 with the totally unnecessary addition - also to be found quite without justification in the IBM z/OS Communication Server TN3270E server implementation - of "3279" as well as "3278" - utterly unnecessary and utterly a waste of time and effort - see "7.1 DEVICE-TYPE Negotiation" in the RFC.[1]
 
-
 
A few comments on your definitions - in case anyone is tempted to copy them without checking on whether they are "in proper shape".
 
The VBUILD operands MAXGRP, MAXNO and MAXDLUR have been obsolete for a decade or so now. The same is true of the PU statement MAXPATH operand.
 
Given that the LU statements are now selected by the DDDLUS function, I don't believe the VPACING operand, an LU statement operand for SSCP-dependent LU statements, can be relied upon in the same way as it would for LU statements actually following the PU statement. In any case, in this configuration there is no session stage over the subarea network between boundary functions which would call on the pacing to which the VPACING operand value refers.
 
The ADDR operand is irrelevant in this context and is mistakenly described in the manuals as required. I persuaded VTAM development to make it an optional operand about 15 years ago but, although the change was made, it never made it through to the "Req" column in the manuals.
 
In this context, a 3172 and, indeed, a 3172 simulated internally inside CIP, I don't expect the MAXDATA is needed.
 
MAXOUT is relevant only for SDLC non-switched lines. At least you haven't specified PASSLIM or IRETRY which I so often see on these sorts of definitions, probably ported in antiquity from non-switched SDLC lines.
 
You never need to specify MODETAB=ISTINCLM. It's not really a default value for the MODETAB operand since whatever is specified for the MODETAB operand is, in effect, concatenated to mode table ISTINCLM. Thus, if the MODETAB operand is not specified, there is, actually no default name value necessary. ISTINCLM is always available.
 
And just now I recall we are in the VSE list where - I thought - the IBM-supplied mode table had a name other than ISTINCLM - but I could be wrong ...
 
-
 
Anyhow, OSA features - defined with CHPID type OSE - could use these VTAM definitions but only if you had some "outboard" implementation of TN3270(E) other than the CIP.
 
Chris Mason
 
[1] Even if the "3279" codes have been covered because those codes are mentioned in the predecessor RFCs 1576 and 1340, those RFCs are quite clear concerning the 3279 devices that actually were announced in the late '70s, namely the 3279-2 and 3279-3. There never was nor never will be a 3279-4 or 3279-5!!!!!
----- Original Message -----
Sent: Monday, March 15, 2010 2:47 PM
Subject: RE: An OSA TN3270 question

Chris

No sense in guessing, these are the Vtam Definitions used for the CIP.

Chris Mason

unread,
Mar 15, 2010, 12:59:22 PM3/15/10
to
Dave
 
> If you just want the OSA to supply the LAN interface, then that would generally be OSA-2 (OSE in the IOCDS) or OSA-X (OSD in the IOCDS) and would support more like 4096 sessions per OSA port.
 
More "apples and oranges" - and not quite fully accurate.
 
> ... then that would generally be OSA-2 (OSE in the IOCDS) ...
 
By CHPID type OSE, you mean to indicate SNA use of OSA features - although CHPID type OSE also allows IP use - and, by CHPID type OSE, you mean to indicate IP use of OSA features.
 
Here is the table taken from the current Open Systems Adapter-Express Customer’s Guide and Reference:
 
<quote>
 
OSA-Express Physical Feature Descriptions
 
OSA-Express supports various LAN types. The following is a table of the OSA-Express features available.
 
Table 2. OSA-Express Features
 
Feature Name                          Channel Type      Feature Code
 
OSA-Express Gigabit Ethernet LX       OSD               1364
OSA-Express Gigabit Ethernet SX       OSD               1365
OSA-Express 1000Base-T Ethernet       OSE or OSD        1366
OSA-Express Gigabit Ethernet LX       OSD               2364
OSA-Express Gigabit Ethernet SX       OSD               2365
OSA-Express Fast Ethernet             OSE or OSD        2366
OSA-Express Token Ring                OSE or OSD        2367
 
OSA-Express2 Gigabit Ethernet LX      OSD or OSN        3364
OSA-Express2 Gigabit Ethernet SX      OSD or OSN        3365
OSA-Express2 10 Gigabit Ethernet LR   OSD               3368
OSA-Express2 1000Base-T Ethernet      OSE, OSD or OSN   3366
 
OSA-Express3 10 Gigabit Ethernet LR   OSD               3370
OSA-Express3 10 Gigabit Ethernet SR   OSD               3371
OSA-Express3 GbE LX                   OSD               3362
OSA-Express3 GbE SX                   OSD               3363
OSA-Express3 1000Base-T Ethernet      OSE, OSD or OSN   3367
 
OSA-Express3-2P GbE SX                OSD               3373
OSA-Express3-2P 1000Base-T Ethernet   OSE, OSD or OSN   3369
 
Notes:
 
1. Feature codes 1364 and 1365 cannot be ordered after the OSA-Express2 Gigabit Ethernet features are available.
 
2. Feature codes 2364, 2365, 2366 and 2367 can be carried forward on an upgrade from z900 to z990.
 
3. Feature code 2367 is not supported on z9 and above.
 
4. Feature codes 3373 and 3369 are available only on z10 BC.
 
</quote>
 
Thus you can easily see that CHPID - or channel - type OSE is *not* restricted to OSA-2.
 
> ... and would support more like 4096 sessions per OSA port.
 
The first occurrence of the number 4096 in the current Open Systems Adapter-Express Customer’s Guide and Reference is in the section "SNA Modes for Channel Type OSE on z/OS":
 
<quote>
 
4. Specify the maximum number of stations, or PUs, for each port. A maximum of 4096 PUs can be specified for an OSA-Express CHPID. See page 106 and the SNA mode requirements for each operating system in the earlier chapters.
 
</quote>
 
A link station and a potential associated SSCP-dependent PU entity is *not* a "session".
 
> ... or OSA-X (OSD in the IOCDS) and would support more like 4096 sessions per OSA port.
 
Furthermore the 4096 *connection* restriction applies *only* to "connection-oriented" SNA where each SNA connection between adjacent link stations takes up storage in the logic supporting the link station. This restriction doers *not* apply to "connectionless" IP since there is no corresponding demand on storage in the IP interface logic.
 
Chris Mason
----- Original Message -----
Sent: Monday, March 15, 2010 4:01 PM
Subject: Re: An OSA TN3270 question

indust...@winwholesale.com

unread,
Mar 15, 2010, 1:23:49 PM3/15/10
to
owner...@Lehigh.EDU wrote on 03/15/2010 12:58:43 PM:
> Thus you can easily see that CHPID - or channel -  type OSE is *not*
> restricted to OSA-2.


        Didn't say it was -- didn't want to write a book, either.  People generally get lost in book-type responses and generally don't read them anyway.  They also generally end up asking the more pointed questions which a shorter response generally encourages anyway.  I used the word "generally" (both previously and now) to indicate these were/are not hard and fast rules and were/are only how I have successfully dealt with these things in the past.  YMMV

rbo...@aol.com

unread,
Mar 15, 2010, 2:18:09 PM3/15/10
to
I'm looking for someone who has setup the above and would care to share
their setup. I believe I've got the IOCP definitions correct, but if
anyone would like to send theirs anyway, it would be appreciated.

The ONLY IBM Redbook I found relates to z/OS, VM and Linux. Can't find
anything for z/VSE. I also can't find anything for the CSI TCP/IP side.


TIA

Mark Pace

unread,
Mar 15, 2010, 2:28:12 PM3/15/10
to
ADD 724:726,OSAX,01

DEF LINK ID=HIPX01 TY=OSAX DEV=(724,725) DATAP=726 PORTNAME=IT2TA -    
    IP=10.6.0.9                                                        
DEF MASK ID=MSKINT NET=10.6.0.0 MASK=255.255.255.0                                                                      
DEF ROUTE ID=INTNET LINK=HIPX01 IP=10.6.0.0
--
Mark Pace
Mainline Information Systems
1700 Summit Lake Drive
Tallahassee, FL. 32317

Helio Mario Neves Pimentel de Oliveira

unread,
Mar 15, 2010, 3:31:37 PM3/15/10
to
Hi Rbot,

In the $IPLESA member:
000010 800,$$A$SUPX,LOG,NOPDS,VIO=512K,VPOOL=64K
000020 ADD ...
000050 ADD 500:507,CTCA,EML
000070 ADD 100:102,OSAX OSA
000080 ADD 1FE,OSAD
000090 ADD 800:802,3277 OSA-ICC
000100 ADD E00:E03,OSAX,1 HIPERSOCKET
000110 ADD F40,FBAV VIRTUAL DISK - TYPE FBA
000120 ADD F90,3480 VIRTUAL TAPE
000130 ADD FDF,FBAV VIRTUAL DISK - TYPE FBA - STD LABEL
000140 ADD FE1,3800 POWER VIRTUAL PRINTER #1
...

In the CSI INIT member:
* ------------------------------------------------
* IPINIT00.L - ENGEPEL TCP/IP V1.5.F
* Z/890 - ATUALIZADO EM 24/11/2009
* ---------------------------------------
* HOST IBM VIA OSA = 192.168.000.009 OSA.VSE.EGP
* HOST IBM VIA HIPERSOCKET = 192.168.006.009 HIP.VSE.EGP
* HOST IBM VIA 2216 = 192.168.009.001 2216.VSE.EGP
*
* IBM 2216 PORTA ESCON = 192.168.009.002
* IBM 2216 PORTA ETH 0 = 192.168.040.200 REDE INT EGP USO LOEL
* IBM 2216 PORTA ETH 1 = 192.168.008.001 REDE INT EGP USO OUTROS
* IBM 2216 PORTA ETH 3 = 192.168.000.200 REDE INT EGP
* DNS1 = 192.168.000.253
* DNS2 = 192.168.039.038
*
* ------------------------------------------------
* Define IP address and subnet mask for VSE
* ------------------------------------------------
SET IPADDR = 192.168.009.001
SET MASK = 255.255.255.000
SET DNS1 = 192.168.000.253
SET DNS2 = 192.168.039.038
SET DEFAULT_DOMAIN = egp
GATEWAY OFF
*
* ------------------------------------------------
* Define the Communication Links *
* ------------------------------------------------
DEFINE LINK,ID=IBMOSA,TYPE=OSAX,DEV=100,DATAPATH=102,FORCE,MTU=1492, -
PORTNAME=OSAXPORT,IPADDR=192.168.0.9
DEFINE LINK,ID=IBMHIP,TYPE=OSAX,DEV=E00,DATAPATH=E02,FORCE,MTU=32768, -
IPADDR=192.168.6.9
DEFINE LINK,ID=IBM2216,TYPE=3172,DEV=500,FORCE,MTU=1500, -
IPADDR=192.168.9.1
DEFINE ADAPTER,LINKID=IBM2216,TYPE=ETHERNET
*
*---------------------------------------------------------------------*
* Define Routing Table *
* DYNAMIC_ROUTE should always be turned OFF once your *
* routing table has been properly configured. *
*---------------------------------------------------------------------*
DYNAMIC_ROUTE OFF
DEFINE ROUTE,ID=ROTA_HIP,LINKID=IBMHIP,IPADDR=192.168.6.0
DEFINE ROUTE,ID=GTWY_NET,LINKID=IBM2216,IPADDR=0.0.0.0, -
GATEWAY=192.168.009.002,ADAPTER=0
...


Helio Mario - Brasil


-----Mensagem original-----
De: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] Em nome de
rbo...@aol.com
Enviada em: Monday, March 15, 2010 3:17 PM
Para: VSE Discussion List
Assunto: HiperSockets

Michael Simms

unread,
Mar 15, 2010, 3:31:55 PM3/15/10
to
Don't mean to spill any more complications into the mix but what if.....
We had zVM. We had 'zLinux'.
Could we use VSwitch in the mix here some how?
Only problem is we don't have VM/VTAM. This may not be needed anyway.
But still?
Ideas?

Thanks!!


From: "indust...@winwholesale.com" <indust...@winwholesale.com>
To: VSE Discussion List <vs...@Lehigh.EDU>
Sent: Mon, March 15, 2010 1:22:59 PM

Subject: Re: An OSA TN3270 question

Frank M. Ramaekers

unread,
Mar 15, 2010, 4:26:21 PM3/15/10
to

Yes with z/VM you can use VSwitch for both your z/VSE guests and z/Linux guests.

 

(VM/VTAM not required.)

 

 

Frank M. Ramaekers Jr.

 

 


From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Michael Simms


Sent: Monday, March 15, 2010 2:31 PM
To: VSE Discussion List

_____________________________________________________ This message contains information which is privileged and confidential and is solely for the use of the intended recipient. If you are not the intended recipient, be aware that any review, disclosure, copying, distribution, or use of the contents of this message is strictly prohibited. If you have received this in error, please destroy it immediately and notify us at Priva...@ailife.com.

Helio Mario Neves Pimentel de Oliveira

unread,
Mar 15, 2010, 9:25:46 PM3/15/10
to
And now the IOCP I forgot:

...
* =================================================================== *
* ***** HiperSockets
* =================================================================== *
CHPID PATH=F0,TYPE=IQD,CHPARM=80,SHARED
*
CNTLUNIT CUNUMBR=F000,PATH=(F0),UNIT=IQD
IODEVICE ADDRESS=(0E00,3),CUNUMBR=(F000),UNIT=IQD,UNITADD=00

Helio Mario


-----Mensagem original-----
De: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] Em nome de Helio
Mario Neves Pimentel de Oliveira
Enviada em: Monday, March 15, 2010 4:33 PM
Para: VSE Discussion List
Assunto: RES: HiperSockets

Thomas Hebert

unread,
Mar 15, 2010, 9:44:17 PM3/15/10
to

Frank is correct. Here are a few additional pointers:

The VSWITCH will pass all trafic from the virtual NICs to the LOCAL LAN segment
that the real hardware is plugged into. This means that each TCPIP stack that uses the switch will get it's own IP address, etc. VTAM is not needed. Note that I show you
below how to configure VM's TCPIP stack to use the VSWITCH. You don't have to bring
up TCPIP on VM if you only want to use the switch with z/OS or z/VSE. You can't share the OSA adapter directly accross VMs. So you must dedicate the real OSA adapter to the VSWITCH. To keep using VM's TCPIP you must configure it to use the switch as well.


1: We added the following line to AUTOLOG1'S PROFILE EXEC. 420 is the real address of our OSA. Ours is a zPDT awsosa device. Also Grant access to TCPIP and to the other VMs that will use it.

cp define vswitch vsw1 rdev 420 connect
cp set vswitch vsw1 grant tcpip
cp set vswitch vsw1 grant zos1
cp set vswitch vsw1 grant vse1

2: In TCPIP's USER DIRECT entry, we add this NICDEF. This is a virtual OSA QDIO adapter, at address 400, that is virtually plugged into the VSWITCH named vsw1.

NICDEF 400 TYPE QDIO DEV 3 LAN SYSTEM VSW1


3: VM's PROFILE TCPIP has this link telling TCPIP that a OSA/QDIO device is available at address 400.

DEVICE OSAQDIO1 OSD 400 PORTNAME OSA1 PRIROUTER AUTORESTART
LINK ETH0 QDIOETHERNET OSAQDIO1

4: eth0 is our only adapter so all our network definitions reference it.

5: We issue this at the end of the profile

START OSAQDIO1

This brings starts up the virtual adapter when TCPIP comes up on VM. Follow IBM's docs to get TN3270 working.


6: We then added this line to each VM that wants to use the virtual switch

NICDEF 120 TYPE QDIO DEV 3 LAN SYSTEM VSW1

7: for each VM configure the guest TCPIP stack to use a QDIO capable adapter at the address specified in 6:

Let me know if this works for you.

Cheers,

Tom Hebert


________________________________
> Date: Mon, 15 Mar 2010 15:25:40 -0500
> From: FRama...@ailife.com
> To: vs...@Lehigh.EDU

> Subject: RE: An OSA TN3270 question
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Yes with z/VM you can use VSwitch for both
> your z/VSE guests and z/Linux guests.
>
>
>
>
>
>
>
> (VM/VTAM not required.)
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Frank M. Ramaekers Jr.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> ________________________________
>
>
>
>
>
> From:
> owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf Of Michael Simms
>
> Sent: Monday, March 15, 2010 2:31
> PM
>
> To: VSE Discussion List
>
> Subject: Re: An OSA TN3270
> question
>
>
>
>
>
>
>
>
>
>
>
>
>
> Don't mean to spill
> any more complications into the mix but what if.....
>
> We had zVM. We had 'zLinux'.
>
> Could we use VSwitch in the mix here some how?
>
> Only problem is we don't have VM/VTAM. This may not be needed anyway.
>
> But still?
>
> Ideas?
>
>
>
> Thanks!!
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> ________________________________
>
>
>
>
>
> From:
> "indust...@winwholesale.com"
>
> To: VSE Discussion List

Chris Mason

unread,
Mar 16, 2010, 7:06:18 AM3/16/10
to
Dave
 
Testy is it? Don't like errors being pointed out! I can do testy with the best.
 
> Didn't say it was ...
> I used the word "generally" ...
 
I assessed the "generally" but decided that the overall impression that would be left to the casual - dare I say impressionable? - reader/lurker would be that you would "generally" use OSA-2 for CHPID type OSE and SNA whereas, as my extract from the "book" indicated, the very latest and best OSA supports OSE:
 
<quote> OSA-Express3-2P 1000Base-T Ethernet   OSE, OSD or OSN   3369 </quote>
 
So there's no need to worry about whether SNA will be supported into the future - to bring out what was hidden between the lines - intentionally or not I couldn't say.
 
> ... -- didn't want to write a book, either.
 
This aversion to "books", actually "reading" rather than "writing" is a general (!) problem - some might say "issue" - with list responses - I won't say specifically the VSE list. A small check with the main OSA manual would have revealed both errors regarding your 4,096 - which, TGCWCID, was a handy point to raise since the *capacity* of any solution had been emphasised. As it happens Sid needs only 6 out of the 4,096 SNA *connections*.
 
Discovering that the number appeared in a section discussing SNA with CHPID type OSE *might* have led to a realisation that such a restriction would apply only to SNA with CHPID type OSE and *not* CHPID type OSD.
 
"People get lost" with responses like the rest of this response - at least this person does!
 
Chris Mason
----- Original Message -----
Sent: Monday, March 15, 2010 6:22 PM
Subject: Re: An OSA TN3270 question

Chris Mason

unread,
Mar 16, 2010, 7:08:56 AM3/16/10
to
Michael
 
The problem this thread is addressing is how to set up a DR site *without* a CIP providing a TN3270[1] server function where the site being "copied" has such a service with attendant VTAM definitions supporting CICS.
 
There was a suggestion that CICS might be accessed directly from an IP-based software layer. Unfortunately I have not been keeping current with recent CICS developments but I believe that CICS transactions using IP-based software layers need to be rewritten from a comparable transaction using the traditional VTAM support. I'm sure I will be jumped on promptly if this is not the case.
 
Such rewriting seems incompatible with the requirements of setting up a DR site.
 
In this context, only a restatement of the original post - as I understood it, what is the relevance of your post? - or do we have some sort of tangential topic which seems to have as its theme how to do without VTAM!!!
 
Of course, if there is a way to access CICS transactions with an intermediate TN3270 server without the agency of VTAM, I'd like to hear about it - even if I can just about imagine how it could be done - as long as you didn't want to present it as a technique oriented to "business" applications.
 
Chris Mason
 
[1] Probably specifically a TN3270*E* server.
----- Original Message -----

Edward M Martin

unread,
Mar 16, 2010, 9:50:56 AM3/16/10
to

Hello Sid,

 

     Not much at all.  Typical from TMON VSE ACTIVITY screen on the PROD system is 4-5%,  Operators have indicated they have never seen it higher than 7%.

Tobias, Sid

unread,
Mar 16, 2010, 9:56:48 AM3/16/10
to

Edward

 

That’s a lot less than I thought it would be.

 

Thank you

 

Sid Tobias

Health Management Associates, Inc.

Manager, Computer Operations

239 552 3462

Edward M Martin

unread,
Mar 16, 2010, 10:15:30 AM3/16/10
to

Hello Sid,

 

Yes it is a small amount.  The big thing is storage.  We use POOL=NO on all our TELNETD’s  and

On the GPSD we are using storage in our PRODLIB.AULT (TESTLIB and BETALIB in their respective systems).

 

The TELNETD/GPSD partition is 70 meg. 

 

Finally,  We have found that starting the GPSD stuff after the partition is up seems to help with the storage allocations.

indust...@winwholesale.com

unread,
Mar 16, 2010, 11:19:05 AM3/16/10
to
owner...@Lehigh.EDU wrote on 03/16/2010 07:05:29 AM:
> Don't like errors being pointed out!


        If I were trying to quote facts, then I wouldn't mind being corrected -- in fact, I welcome it.  There are many on this list who know how to do that with tact.  I just don't like those whom choose to read things into what other people say and then erroneously use it as a soapbox on which to blow their own horn.  In case it has never dawned on you, if you have the urge to blow your own horn, it is very easy to do so without (erroneously) having to step on other people in the process.  In other words, if it still isn't clear to you, you can spout all the facts and detailed information you like without even mentioning/quoting prior posts or individual names.  I can demonstrate that for you if you need a guide.

Sincerely,

Dave Clark

WinWholesale Group Services
3110 Kettering Boulevard
Dayton, Ohio  45439  USA
(937) 294-5331

Rbo...@aol.com

unread,
Mar 16, 2010, 9:15:54 PM3/16/10
to
Mark and Helio,
 
 
Thank you for the stuff on HiperSockets. I have successfully setup IP PNET for 2 of our 9 LPAR's.
 
I do have a question for you both and any others who care to respond that didn't previously (hard to believe that only 2 sites are using HiperSockets which I guess means that those that didn't are only using 1 LPAR or didn't want to respond in the first place).
 
Other than IP PNET are you using HiperSockets for z/VSE<>z/OS or z/VSE<>Linux or something else?

Mark Pace

unread,
Mar 17, 2010, 8:19:34 AM3/17/10
to
All of the above.  I have hipersocket connectivity on all zSeries OSes  VSE, VM, Z/OS, & Linux.

VM RSCS / VSE PNET / z/OS JES    -   Our Linux does DNS & FTP services.

Helio Mario Neves Pimentel de Oliveira

unread,
Mar 17, 2010, 9:42:06 AM3/17/10
to

We have 3 LPAR’s = VSE & 2 Linux Suse.

We have hipersocket connection  on VSE & Linux.

 

Helio Mario - Brasil

 

 

De: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] Em nome de Mark Pace
Enviada em: Wednesday, March 17, 2010 9:19 AM
Para:

VSE Discussion List
Assunto: Re: RES: HiperSockets

rbo...@aol.com

unread,
Mar 17, 2010, 12:05:29 PM3/17/10
to
Helio,


Are you using IP PNET w/HiperSockets?


-----Original Message-----
From: Helio Mario Neves Pimentel de Oliveira <h...@engepel.com.br>
To: VSE Discussion List <vs...@Lehigh.EDU>

Helio Mario Neves Pimentel de Oliveira

unread,
Mar 17, 2010, 3:05:23 PM3/17/10
to
No, not even PNET.

Helio Mario


-----Mensagem original-----
De: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] Em nome de rbo...@aol.com
Enviada em: Wednesday, March 17, 2010 1:04 PM
Para: VSE Discussion List
Assunto: Re: RES: RES: HiperSockets

Rbo...@aol.com

unread,
Mar 18, 2010, 4:11:08 PM3/18/10
to
Frank-See below.
 
Just curious, have you tried other transports to see if the results are
similar?  (FTP?)
No. I suppose I could FTP the same LST queue entry from one of the HiperSockets LPAR's to the other HiperSockets LPAR. Then do the same to one of the non-HiperSockets LPAR's to the other non-HiperSockets LPAR.


We had the same results (with respect to PNET).  It may be PNET that is
slow.
While it been a number of years I do remember that when I first setup IP PNET it beat the pants off PNET using CTC connections.



Frank M. Ramaekers Jr.




-----Original Message-----
From: owner...@Lehigh.EDU [mailto:owner...@Lehigh.EDU] On Behalf
Of rbo...@aol.com
Sent: Thursday, March 18, 2010 11:52 AM
To: VSE Discussion List
Subject: HiperSockets

Don't mean to beat this to death but thought some of you would like to
know the results of some testing I did.

I wrote 500,000 133 byte records to disk and then used DITTO to print
them. This resulted in 26,316 pages containing 1,026,346 lines. This
was done on one of the LPAR's I was using HiperSockets (i.e. LPAR 3). I
then PNET'ed the LST queue entry to the second LPAR I had setup using
HiperSockets. The wall clock timings I used (I don't know of a better
timing method) was from when I PALTER'ed the LST queue entry, for a
starting point, and the ending time I used was when POWER received the
LST queue entry on the target LPAR (this is indicated my message
1RB5I). Also, I ignored the seconds.

Time I Issued The PALTER     Time LST Que Entry Received             
Total Time

14:05                                            14:24                 
                                         19 minutes  
HiperSocket_To_HiperSocket

I repeated the this using LPAR's that were NOT using HiperSockets

Time I Issued The PALTER     Time LST Que Entry Received             
Total Time

14:46                                            15:04                 
                                         18 minutes  
Non_HiperSocket_To_Non-HiperSocket


I have to admit, I'm NOT impressed with the results. I would've
expected HiperSockets_To_HiperSockets would've been faster.

If anyone has an explaination or has some suggestions to improve the
above, please let me know.


Since using HiperSockets doesn't appear to be much worse than not using
HiperSockets I intend to fully implement this anyway,in early May. For
one thing using HiperSockets does eliminate any possibility of network
issues. It also eliminates in possibly having to change IP address, in
our NDT's, for DR.
0 new messages