Tweet
I need to do a refresh, of data only, from production to test on a regular basis to keep data current. A few of our files are very large with history so only an 1/8 of the records in those files will be pulled over from production to test; otherwise, the files will be the same as prod. Knowing that it cannot be a complete duplicate of what is in production, what is the best method to pull this data over on a weekly basis? Would sndnetf be an option or too clunky? I need a programming method and not a tool I have to purchase. Thanks. Deb
-
-
Re:Refresh from production to test system
I take it that test is another system based on your mention of sndnetf correct? Many ways to do what you need weekly, daily, or even real-time but I'm going to guess the data altogether is large and your developers don't need fresh data from production more frequently. Under the above assumptions I would recommend saving prod files to a *savf which you ftp to the test system and restore from the *savf on the test system. Depending on the size of the *savf and the speed of your connectivity between systems will determine how long it will take to send the data. Since you only wanted data sent, you would have to restore to a temporary library and presumably run CPYF from each physical to the target physicals and probably rebuilding indexes as you go. Perhaps it would be better to do a complete save of data and access paths. When restored on the test system you could delete 72% of your large history files to reduce storage. Basic OS commands would do it all. It would take a little programming effort to do this "automatically". Maybe do it interactively first and later put each command you used in a CLLE program job scheduled to run once a week...??? Hope this helps. Tell us how you do? -
Re:Refresh from production to test system
I don't know where I got 72% from...??? I know you know.Comment
Comment