Imagine having multiple projects working on your BI system concurrently. And try to consider the business release cycle that allows transports to be imported to regression and production server based on a straight timeline. Now, what happen to shared objects (esp infoobjects and datasource) being overwritten and missing manual steps before a transport went in? Surely some consideration to strategize the transports have to be done. And this is what the post is about.
Action for Pre cutover:
1) Engagement with basis on system refresh for regression server,logical system name mapping,
RFC connection
2) Engagement with different projects on the manual steps in between transport and data
loading sequence (two projects working on different solution may shared same datasource,
eg.Marketing and Finance retrieve data from COPA)
3) Preparation of transport build recipe
4) Communication to respective stakeholder on dates,system lockout and date for last transport
to be included in build recipe
5) List of users remain unlock and process chain to be 'paused' during cutover
6) Manual steps in between of cutover and party involves. Eg. replication from R/3
Action for Post cutover:
1) Compare inactive objects before and after cutover
2) Ensure all process chain run successfully
3) Reports can be executed successfully
4) Document down lesson learned. Eg. Cube content be truncated prior to sending in a change to
add navigational attribute in order to shorten the transport time
5) Raise awareness to respective party on their changes in their system that impact BI. Eg is
changing logical system name in source system might resulted in delta extraction failure in BI.
Another example is COPA realignment in R/3 will break the delta extraction in BI.
Where is the rest of the post?
ReplyDeleteRegards,