On 12 Nov 2006 18:14:20 -0800, in bit.listserv.ibm-main you wrote: >> much snipped > >These were just some thoughts as I sit here waiting for an IPL to >install new maintenance. I must "do my magic" after the production >system (we only have one) is down. I'm using our "sandbox" to "do the >magic" because I couldn't afford the DASD to duplicate the RES volumes. >Luckily, it is only renaming 4 "minor" datasets (LINKLIB, NUCLEUS, >LPALIB, and MACLIB). I did the maintenance the "bad way". I created >duplicates on the same volumes, with new names and put the maintenance >into them. Yes, this is stupid. No, I didn't really have a choice. We >did have a plus of DASD a few months ago. Then somebody decided that we >needed a new type of "model office" environment and ate almost all the >unused volumes. These same somebodys also want a new set of test volumes >for a alternate test environment. Then they want yet another environment >called "regression". That is 6 environments: Prod, MDOF1, MDOF2, Test1, >Test2, and REGRESSION. And our shop is not that large! Like 50 MSUs >total (prox 300 MIPS - the evil word) on a z890.
As someone who has done both applications and systems programming, it can be a nightmare getting coordinated test data. If you have a number of things going on, DASD can get eaten quickly. Data base is worse than VSAM because you can't mix and match such as use a test invoice table but the production customer and product tables in a read only manner. Then if you are doing a noticeable change, you may want to have some pathological test data to exercise the error paths. Sometimes the simplest thing to do is full volume testing because you can use point in time backups to start the process. Now with privacy laws and other legislative constraints, testing may get even more interesting and designing masking systems to sanitize the data while still having it usable may be very challenging in some cases. ---------------------------------------------------------------------- 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
