On Thu, 21 May 2009 07:53:08 -0500, McKown, John <[email protected]> wrote:
>> -----Original Message----- >> From: IBM Mainframe Discussion List >> [mailto:[email protected]] On Behalf Of Mark Wilson >> Sent: Thursday, May 21, 2009 1:30 AM >> To: [email protected] >> Subject: PDSMAN Replacement >> >> Hi, >> >> I have the task of removing CA-PDSMAN from two of our systems. >> >> We currently use the following functions: >> >> * PDSMAN Library Space Reuse facility >> * Dynamic BLDL Facility >> * Remote Data Compare >> >> I believe I can convert my PDS¹s to PDS/e¹s to negate the >> need for the Space >> Reuse facility. > >Likely true. However, remember that a PDSE cannot be shared except within a single SYSPLEX. If you try to share it across multiple SYSPLEXes or MONOPLEXes, then "results are undefined" and likely to be terrible (corrupted PDSEs). > >> >> The Dynamic BLDL function causes me a little concern. I have >> inherited a >> situation where we have one massive load library that >> contains all batch & >> online application programs. All batch jobs and online >> systems have either a >> JOBLIB or STEPLIB DD Statement that references this one >> library. Many years >> ago (well before my time) the techies turned off this PDSMAN >> feature and had >> very severe system performance issues for online and batch >> applications, >> resulting in this feature being re-enabled. I was think of >> splitting the >> library, perhaps by application to reduce the search time for none >> application code. > >Use LLA. It can be used to buffer any load library, not just the LINKLIST. > >./ ADD NAME=CSVLLAnn >LIBRARIES(-LNKLST-, > NPHPN.PGM.LOADLIB, > DCHPN.ENDEVOR.EZTPLUS.LOADLIB) >FREEZE(-LNKLST-, > NPHPN.PGM.LOADLIB, > DCHPN.ENDEVOR.EZTPLUS.LOADLIB) >./ENDUP > >is an example. > Since the OP said he would convert to PDSE, he can take advantage PDSE / SMS caching (so make sure the PDSE is SMS controlled). Be sure to review technote "TIPS0567 - Partitioned Data Set Extended (PDSE) Hiperspace (Caching) Problems, Considerations, and Parameter Changes" which explains the caveats including CPU cycles (I have no idea how much PDSMAN Dynamic BLDL Facility "costs" in terms of cycles and storage). In addition the OP should make sure to use the IGDSMSxx parm / setting: PDSE1_BUFFER_BEYOND_CLOSE(YES) PDSE_BUFFER_BEYOND_CLOSE(YES) isn't really helpful since the LNKLST is never closed anyway. (note the difference between PDSE1_ & PDSE_) The Partitioned Data Set Extended Usage Guide Redbook is a good source also (SG24-6106). If you search the Redbook site you should be able to find both of the above. >> >> The remote Data Compare is used in batch to compare PDS¹s on the QA & >> Production systems. A checksum is generated for each PDS and >> then compared. >> If a difference is found a WTO is posted (picked up by >> automation) that >> triggers a check of the libraries. > >I can't help with this one. > Nothing slick. Perhaps if you can get a shared volume between the 2 environments you can run a job that copies one of the 2 libraries you are comparing in the first step to the shared volume and then submits a job (NJE, FTP) to run on the other system that uses IEBCOMPR. The 2nd step of the submitted job could use any standard "opswto" program to generate the WTO / WTOR based on a non-zero return code. Mark -- Mark Zelden Sr. Software and Systems Architect - z/OS Team Lead Zurich North America / Farmers Insurance Group - ZFUS G-ITO mailto:[email protected] z/OS Systems Programming expert at http://expertanswercenter.techtarget.com/ Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.html ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
