Logical Files and UAT deployments
Posted on 2012-03-22
We have a so called "staging" AS400 server which gets periodically backed up with data from production. This server is used for various business purposes including preparing for UAT deployments. Before a UAT deployment, these are the steps that we perform on this server:
1) Create duplicates for all source, objects and data files (not database tables).
2) For purposes of security, we scramble and subset the data in the files.
3) All this is then put into a save file and pushed out to the respective UAT servers.
4) A remote restore save file command is issued on the staging server which deploys the code on the UAT server.
The problem i am facing is in step# 2 above. An elaboration of step# 2 is as follows:
1) After a duplicate of production data files are created, I delete all LF's that exist on the PF's.
2) Remove all members from all PF's.
3) Based on our business needs, I add back a version (or not) with specific member names.
4) Scramble and subset that single member if needed.
5) Add back all the LF's that were earlier deleted.
All the above steps work fine except that some of the LF's that we have in our libraries are based on 1-2 tables in a schema. Ideally, when I am re-creating the LF, I want the LF to be recreated using the schema name that I am going to use on the UAT server. Most of the times, we do not have that schema name present on the "staging" server. I either end up not creating that LF or if it creates, it is pointed to one of the schema names that exist on staging.
Is there any work-around?
My restrictions are:
1) I have to run all these steps on the "staging" server.
2) We have multiple UAT regions, so it is not feasible to have a program deployed on all UAT servers to take care of this problem with the LF.
3) I want to avoid hard coding of any kind because this program would be used across applications.
The version of OS we are currently on is V5R4 (with plans to migrate to V7R1 this summer).