Dear all 

  The summary is very very clear. Thanks you all for your help!

 I want to ask you other questions:

 >> when considering DFSMS compression , one should take into account that
 >>it comes with a CPU price tag, so I would not compress everything
 >>indiscriminately (unless CPU is cheaper than DASD for me).  


Q1.  How to determine  whether  ext-PS((small and large) and VSAM need be 
compressed or not? how to balance DASD with CPU?
       Is there any tools to analyse and monitor it?

 >> If Ext  P:  Preferred.  The system attempts to allocate the data sets in    
 >>   
 >>   extended sequential format.  If the necessary system resources    
 >>  are not available to allocate them as such, the system allocates  
 >> them as non-extended sequential.                                  

 Q2.What situations is that the necessary system resources 
      are not available to allocate them?

>> Use the EXTENDED ADDRESSABILITY field to provide data sets with
>> addressability of more than 4 gigabytes.                       
 
 Q3. Is EXTENDED ADDRESSABILITY only used for VSAM?

 Q4. Could I assign all VSAM for applications including < 4G and > 4G to use 
EXTENDED ADDRESSABILITY = yes? 

 Q5  Is there any disadvantage for ADDRESSABILITY = yes? 

Thanks a lot!

Best Regards,

Jason Cai 





发件人: R.S. 
发送时间: 2011-04-21  03:31:38 
收件人: IBM-MAIN 
抄送: 
主题: Re: Is DSNTYPE=EXT parameter only used for VSAM 
 
Gentlemen,
Please, re-read carefully, all of us are right here :-)
1. I was right saying that dss output cannot be ext-PS. This restriction 
is removed in z/OS 1.12 - but still true in older versions of z/OS, not 
to mention that at least one of them is still supported. So, the 
restriction "is", not "was" ;-)))
2. Ifrat was right, because he mentioned about the restriction relief.
3. Frank is also right, because he mentioned still valid restriction.
4. Jonh is right, because he mentioned "other ised" of dss - ability to 
dump/restore EF datasets, not the dump dataset characteristics.
5. Frank is right again, beacuse his z/OS is 1.10, so the restriction is 
still valid.
And all of us (but not Yifrat) is a little bit not right...
Here in Poland we call it Czech movie, but the meaning is close to 
Comedy of Errors (as far as I remember the author wasn't Czech).
To summarize:
Dss *output* file can be ext-PS in z/OS 1.12, but it couldn't be in 
previous releases of z/OS.
Dss intput files can be extented format, it have been able to for many 
years and z/OS releases.
Regards
-- 
Radoslaw Skorupka
Lodz, Poland
--

Reply via email to