Problem is usually defined as something is broken. So it is not a problem that a huge pre-built image is loaded when you might use less than 2% of it. You can always buy more memory to support all the waste that takes place for non- problems. So this is purely an option for those shops that do not want to waste all that memory by loading conversion techniques they will never use.
Letting DB2 cause the pre-built image to be loaded did cause a problem if more then one DB2 was started in a short space of time, so a PTF was made available. I'll stick with a CUNUNI00 member loading one always used conversion technique and let z/OS load others. I did hit a problem where an off-the-shelf vendor application running in CICS has a problem getting the conversion techniques they need loaded. Because of time constraints and an inability to recreate the problem outside of their application I ended up running a PLT program at startup time that performed the same instructions as their failing program, but only those, and they load just fine and then the application works. Is having a CUNUNI00 member really confusing to people? Or has IBM made it confusing because the DB2 people are not on the same page as the UNICODE folks who know better? Shouldn't we unconfuse the DB2 people? And the IBMer who suggested letting DB2 consume the machine? I have always believed if a recommendation helps waste resources then they should be provided for free. On Thu, 3 Jul 2008 19:04:10 -0500, Mark Zelden <[EMAIL PROTECTED]> wrote: >On Thu, 3 Jul 2008 16:52:25 -0500, Brian Peterson ><[EMAIL PROTECTED]> wrote: > >>Well, if you *don't* initialize Unicode On Demand by issuing a simple command >>at IPL time, DB2 Version 8 will detect that Unicode has not been >initialized and >>will attempt to load the huge pre-built image. > >Again, I don't see this as a problem. Apparently either does IBM since the >migration guide recommends removing CUN* members from parmlib. > ---------------------------------------------------------------------- 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
