The Unisoft CELLS Workflow MES interface uses ODBC (Open Database Connectivity). ODBC is a broadly accepted, vendor-neutral mechanism for accessing relational databases. Any ODBC-compliant application can access any Database Management System (DBMS) for which a driver is installed. Drivers are available for all major DBMSs and even for text or CSV files. Consequently, CELLS Workflow can communicate with Microsoft Access, SQL Server, Oracle, Sybase, and even Excel. Both Oracle and Microsoft SQL Server are compatible with CELLS, although they offer more capabilities than what the CELLS product typically requires. Only a minimum configuration of Oracle is necessary.
Generally speaking, there are two steps in getting CELLS to work with a database. The first is to create the tables in the database. The second is to configure ODBC.
Creating the tables in the database is mostly a manual process. It is described in the Appendix 1 manual. A critical action is to define the primary keys for each table. Sometimes the database has a feature that allows tables to be imported from a different database. SQL Server, for example, allows an Access database, like cells32.mdb, to be imported. It is still necessary, however, to explicitly define the primary key for each table.
The second step, configuring ODBC to let CELLS know where the database is located and is done through Control Panel "ODBC Data Source Administrator" or with the program odbcad32.
Microsoft SQL Server has a very flexible facility for copying tables from all sorts of other systems, including Access. Nevertheless, when it copies the tables, it often neglects to copy the primary keys. Therefore, one has to define the primary keys for each table in SQL Server itself.
An example is one could import the sample CELLS.MDB Access database into SQL Server. The import does not copy keys/indexes; therefore, after the import you must manually define primary and secondary indexes in SQL Server.
