On Thu, 14 Aug 2014 21:37:27 -0400, Thomas Conley wrote:
>    ....
>>> I'm looking for ways to implement multiple sessions, if there are any.
>> 
It's time IBM recognized the need and provided a solution so customers
can stop wasting resources struggling to devise circumventions of 20th
Century restrictions imposed by no other vendor of which I know.

>> You'll have to figure a way to eliminate IKJ56425I and IKJ606I. AFAIK, it is 
>> just not possible. If I remember correctly, it is about the name in the TSO 
>> session address space and VTAM assignment of LU against a TSO name. VTAM can 
>> handle multiple *LOGON*, but after successfull logon it can't handle a 
>> second or more address space with the same name within a LPAR.
> 
Generate unique names as needed.  Under Unix System Services, for example,
I can issue 3 times in quick succession the command:

    ( rexx "address TSO; time; address SYSCALL sleep 33; time" ) &

SDSF shows me:

 SDSF DA MVS3     MVS3     PAG  0  CPU/L    11/ 11      LINE 1-11 (11)
 PREFIX=*  DEST=(ALL)  OWNER=User  SORT=JOBNAME/A  SYSNAME=
 NP   JOBNAME   JobID    Owner    C Pos DP Real Paging    SIO   CPU% ASID ASIDX
      User      STC02381 User       LO  FF  386   0.00   3.17   0.27   99 0063
      User      TSU02388 User       IN  F2 1207   0.00   0.00   0.21  101 0065
      User1     STC02453 User       LO  FF  391   0.00   6.86   0.18  161 00A1
      User1     STC02384 User       LO  FF  426   0.00   5.83   0.25  126 007E
      User1     STC02449 User       LO  FF  371   0.00   0.09   0.06  157 009D
      User2     STC02450 User       LO  FF  391   0.00   6.86   0.16  158 009E
      User2     STC02454 User       LO  FF  426   0.00   5.75   0.24  162 00A2
      User3     STC02451 User       LO  FF  426   0.00   5.83   0.23  159 009F
      User8     STC02385 User       LO  FF  371   0.00   0.09   0.06   84 0054
      User9     STC02452 User       LO  FF  371   0.00   0.09   0.07  160 00A0
      User9     STC02382 User       LO  FF  401   0.00   6.86   0.17  122 007A

I'm running 4 instances of the TMP on a single LPAR.  z/OS has generated
unique names where needed.  Only one has a 3270 attached.  Why can't
the same technique be used for sessions started from 3270s?

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to