m-Fin ACCOUNT NUMBER Report

Select By OrgNode

An important new feature introduced with this report is support for OrgNodes.  Where the other m-Fin reports accept only an org code, this report accepts either an org code or an org node.  We hope to bring this feature to the other reports in a subsequent m-Fin release.

Tuning

“Tuning” is improving the run-time performance of a process.  I’ve been working on tuning the new m-Fin ACCOUNT NUMBER report in the Cognos QA environment.  A common use for this report will be to run for ranges of accounts in an org or a modest orgnode, but it is useful also for specific accounts or uncommon account ranges in high level orgnodes such as U0001 — DENVER & ANSCHUTZ MED CAMPUS or D0015 – D-EDUC-DEAN OF EDUCATION.  For example, you might run for just account 487602 in D0015.  For the former kind of run, an org index is fastest.  For the later, however, an account index is fastest.  One tuning approach is to get the database to recognize the best index for the values submitted.  Another is to get the database to use both of the indexes together.  Then it doesn’t have to decide which is better.  For m-Fin ACCOUNT NUMBER I have settled on a technique named star transformation that uses this second approach.  It yields good performance for most runs I have tried.

So when we migrate to production this new report (and the other m-Fin 1.5 reports) this Friday evening, we will also migrate some database changes to the CIW production database at the same time, to enable star transformation for m-Fin ACCOUNT NUMBER.

Tuning is behind-the-scenes, under-the-hood work.  Star transformation isn’t something that you will see or control.  I just want you to know I am paying attention to performance problems and doing my best to remedy them.