I am not the expert in this area, but can relate what I know. We were a very early adopter of Oracle on mainframe and UNIX.
We run both Oracle and DB2 on the mainframe and have many UNIX servers running Oracle. For years, the only way to connect to the DB2 data bases on the mainframe from Oracle on UNIX was to use the Oracle Gateway that we installed on the mainframe with Oracle. Without Oracle, the only recourse was extract, transform, and load as a means of exchanging data with distributed. Then came DB2 UDB. IBM's solution was that if you wanted to connect Oracle/UNIX to DB2 on the mainframe, you had to set up a UNIX or Windows DB2 UDB and manage connections from there. Originally I believe the UDB to MF had to be LU6.2, (yuck!). The UDB had to be used for JDBC and ODBC connections from distributed. We are DB2 7.1, and have used third party products on the MF to provide JDBC and ODBC connectivity directly to DB2 for years. If you want to connect to Oracle on the mainframe, UNIX, or Windows it is pretty much all the same. I know there are several changes that have to be made to SQL statements when they are ported from Oracle to DB2, so the SQL is not exactly the same. Oracle is pretty much Oracle where ever it is, but expect a systems programmer to have to work with the Oracle folks to install it on the mainframe. The installation is not trivial, and in addition if you require connectivity to CICS, SAS, Easytrieve, etc to access Oracle (i.e. which gives you access to your data regardless of whether the database is on the mainframe or other platforms) there is extra work and support needed. Some of the above product connections may require additional Oracle type licenses that you may not own as a site license. Regards, Ken ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
