cics 
Author Message
 cics

Quote:

>how do you take care of lou(logical unit of work) in cics?

This is starting to sound like a homework assignment.

DD



Tue, 25 Jul 2000 03:00:00 GMT  
 cics

Quote:

> how do you take care of lou(logical unit of work) in cics?

> khyati

I don't understand.  Please rephrase your question in the form of a
question.  What do you mean by 'take care of'?


Tue, 25 Jul 2000 03:00:00 GMT  
 cics


Quote:

>>how do you take care of lou(logical unit of work) in cics?

>This is starting to sound like a homework assignment.

>DD

LOLOL.... Thank you, DD...
I was thinking the same thing....  :-)    Maybe the next question will
be "How do you differentiate between IMS and CICS for optimal
transaction design?"   LOL


____________________________________



Wed, 26 Jul 2000 03:00:00 GMT  
 cics

Quote:

> > > how do you take care of lou(logical unit of work) in cics?

> > > khyati

> > I don't understand.  Please rephrase your question in the form of a
> > question.  What do you mean by 'take care of'?

> the question which i had mentioned is asked in this manner only. okay can
> u just tell me what is logical unit of work?

Well, it's a very complex question now that I understand what you're
asking.  Simply speaking, however, an LUW is the processing that occurs
between syncpoints of a CICS transaction.  Keep in mind that a single
CICS transaction may run transparently on multiple regions and on
multiple platforms depending on the complexity of the transaction and
system setup.  You might want to read the chapter on "Two-phase commit"
(wherever that is) and heed the advice of the other newsgroup posting
about SYNCPOINT, ROLLBACK, etc.  Things get hairy if you are using a
database because there may be SQL commits (EXEC SQL COMMIT END-EXEC)
that are different from your CICS commits (SYNCPOINTs).  There is always
an implicit SYNCPOINT at the end of a transaction (which in
pseudo-conversational lingo is after the presentation of each 'screen'
or map).  Conversational transactions, {*filter*} as they are, usually do
explicit SYNCPOINTs as required (although there aren't too many of these
left these days).  Does this help?


Wed, 26 Jul 2000 03:00:00 GMT  
 cics

Quote:



> >>how do you take care of lou(logical unit of work) in cics?

> >This is starting to sound like a homework assignment.

> >DD

> LOLOL.... Thank you, DD...
> I was thinking the same thing....  :-)    Maybe the next question will
> be "How do you differentiate between IMS and CICS for optimal
> transaction design?"   LOL

Oh no, *much* too complex... these are Standard Interview Questions,
right up there with 'what is the difference between an index and a
subscript?'.  His later comment of

Quote:
> the question which i had mentioned is asked in this manner only. okay can
> u just tell me what is logical unit of work?

... shows that he knows nothing and is merely parroting... not worth the
digital ink.

... then again... maybe it is time for a bit of {*filter*}.

DD



Wed, 26 Jul 2000 03:00:00 GMT  
 cics

Quote:




> > > how do you take care of lou(logical unit of work) in cics?

> > > khyati

> > I don't understand.  Please rephrase your question in the form of a
> > question.  What do you mean by 'take care of'?

> the question which i had mentioned is asked in this manner only. okay can
> u just tell me what is logical unit of work?

Well why didn't you *say* so... the concept of the Logical Unit of Work
dates back to Ancient Greece; it was referred to by Pythogaras and
finally perfected by the Stoics.  The Stoics, known for philosophising
while going shopping (as in 'I thinks I'll get me on down to th' sto'')
were notoriously tight with a drachma; when they paid for philosophising
they wanted to make sure they were getting their money's worth.  After
much debate on what constitutes an appropriate quantity of philosophy.
They looked to the works of their founder, Zeno; unfortunately most of
what he had said was fragmented.  An earnest search revealed a note to
his {*filter*}, referring to a laughable amount of precious bodily fluids
which Zeno had generated during an amorous moment; the major premise was
Zeno's motion, the minor premise was the {*filter*}'s recepetivity thereof
and the conclusion was but a bit of silly jizm.

Now isn't that easy?

DD



Wed, 26 Jul 2000 03:00:00 GMT  
 cics

Quote:


> > > > how do you take care of lou(logical unit of work) in cics?

> > > > khyati

> > > I don't understand.  Please rephrase your question in the form of
> a
> > > question.  What do you mean by 'take care of'?

> > the question which i had mentioned is asked in this manner only.
> okay can
> > u just tell me what is logical unit of work?

> Well, it's a very complex question now that I understand what you're
> asking.  Simply speaking, however, an LUW is the processing that
> occurs
> between syncpoints of a CICS transaction.  Keep in mind that a single
> CICS transaction may run transparently on multiple regions and on
> multiple platforms depending on the complexity of the transaction and
> system setup.  You might want to read the chapter on "Two-phase
> commit"
> (wherever that is) and heed the advice of the other newsgroup posting
> about SYNCPOINT, ROLLBACK, etc.  Things get hairy if you are using a
> database because there may be SQL commits (EXEC SQL COMMIT END-EXEC)
> that are different from your CICS commits (SYNCPOINTs).  There is
> always
> an implicit SYNCPOINT at the end of a transaction (which in
> pseudo-conversational lingo is after the presentation of each 'screen'

> or map).  Conversational transactions, {*filter*} as they are, usually do
> explicit SYNCPOINTs as required (although there aren't too many of
> these
> left these days).  Does this help?

   Sounds right to me. ROFL.

   Richard Jackson
   Houston



Wed, 26 Jul 2000 03:00:00 GMT  
 cics

(snip)

Quote:
> Well, it's a very complex question now that I understand what you're
> asking.  Simply speaking, however, an LUW is the processing that occurs
> between syncpoints of a CICS transaction.  Keep in mind that a single
> CICS transaction may run transparently on multiple regions and on
> multiple platforms depending on the complexity of the transaction and
> system setup.  You might want to read the chapter on "Two-phase commit"
> (wherever that is) and heed the advice of the other newsgroup posting
> about SYNCPOINT, ROLLBACK, etc.  Things get hairy if you are using a
> database because there may be SQL commits (EXEC SQL COMMIT END-EXEC)
> that are different from your CICS commits (SYNCPOINTs).  There is always
> an implicit SYNCPOINT at the end of a transaction (which in
> pseudo-conversational lingo is after the presentation of each 'screen'
> or map).  Conversational transactions, {*filter*} as they are, usually do
> explicit SYNCPOINTs as required (although there aren't too many of these
> left these days).

Very good overview - but you left out the hard {*filter*} & group sex.

Bill {*filter*} <G>

Quote:
>  Does this help?



Thu, 27 Jul 2000 03:00:00 GMT  
 
 [ 20 post ]  Go to page: [1] [2]

 Relevant Pages 

1. Need Example CICS/REXX doing CICS LINK with COMMAREA to non-REXX PGM

2. Follow up: Question for CICS gurus -- CICS CALL statements]

3. Question for CICS gurus -- CICS CALL statements

4. APPC cobol application programming sample (CICS to CICS on different machine)

5. how to communicate with CICS directly nor through CICS Transaction Server for OS/390?

6. How to use a subprogram with cics extension in an non-cics batch program.

7. CICS trace entries

8. SMF 110 recs (CICS stats)

9. description/explanation of CICS macros

10. CICS Systems Programmer Needed in LA

11. MVS/CICS/VTAM training/certification

 

 
Powered by phpBB® Forum Software