Discovery and Development Strategy of BizTalk Migration

With any migration approach, discovery of SeeBeyond code and artifacts is necessary due to the invaluable wealth of information there which has survived due its robustness or been refined to its current state. Not to use it would be an enormous waste. Although a huge amount of it is redundant in BizTalk due to the pre-built infrastructure (ports, adapters, configuration) you can mine a lot of it outright – some of it a simple cut and paste – such as database queries. Hence knowledge of SeeBeyond is inescapable.

To delve into SeeBeyond or to have a meaningful exchange with the SeeBeyond analyst, it would benefit the BizTalk professional greatly if could gain a first look inside SeeBeyond with the concepts and mental models he is already familiar with. Getting sufficient time from the harried, expensive and stretched thin SeeBeyond analyst is likely not going to be easy to come by, so the problem gets worse.

We have painstaking compiled a reference section to help with this comprehension – you will find this in the Appendix A of this paper.

Once the migration is complete, it is necessary to ensure that the new destination data is semantically equivalent to the original destination data (that has been processed with the SeeBeyond implementation). For this to be so, the results of all operations on the new Biztalk solutions must be identical to those on the old SeeBeyond solutions. There is no better way to gain such assurance than to exhaustively perform the same operations on both solutions and to compare the results. We recommend a Cross-Talk approach for verification testing.

In the Cross-Talk QA approach, the data source is flipped from SeeBeyond to BizTalk receivers and the results compared. When verified with a credible set of test queries, a high degree of assurance can be obtained from this approach that the new data store is semantically equivalent to the old one.