Questions about the ERA to Kuali database migration.
It is my understanding that the ERA database is being merged into the Kuali database. This process involves identifying tables with similar purpose and information and then modifying ERA to use the Kuali tables instead of its own. Is this essentially correct?
We are mainly copying the current ER tables into the Kuali database and making minor changes to resolve naming and data type differences. The current integration point between ERA and the current FIS is the data in the FIS_ER schema. These will go away, as we migrate ERA into Kuali and begin to use the operational data from that system.
Where is the documentation of the changes being made?
You're lookin' at it
Are there any cases where a Kuali table has more columns than the equivalent ERA table? Are views being used to trick ERA into continuing to work without modification of source code or is the ERA source being modified? Some of the ERA views use the SQL '*' wildcard in their definitions. This means that a change in the order of columns of an underlying table could cause a silent failure with data being associated with the wrong columns.
ERA source is being super-modified. The integration tables (FIS_ER) are batch fed in from the FIS. Therefore there should not be any instances of the FIS_ER tables having more/less data than what ERA knows about.
Does the Kuali database use referential integrity constraints? If so, will those contraints be spread through the ERA system?
That is currently being discussed by the Kuali architecture group. RI is being handled by OJB.
The automatic copying of data into the existing ERA FIS_ER tables (and the Universal User Table ER.ER_UNVL_USR_T) can be thought of as a functional firewall separating the ERA database from other systems. When the integration is complete, will this separation and replication system remain in place?
No. As ERA (KRA) is integrated into Kuali, this seperation will cease to exist.
