Change Management: Quest Schema Mangler or Embarcadero Change Mangler

We are deploying weekly builds into a production environment with around 1400 database objects. Currently, we "reverse-engineer" INSERT scripts of all the production data (which isn't that much at this point), drop all database objects, and re-create. When the data grows for another month or so, this won't be a practical way to apply changes.

So, we need a tool to identify differences between (let's say) a DEV schema and a PROD schema, and generate the DLL and DML (to modify data for column changes) to apply those changes while preserving the data. Not an easy task!

We initially looked at Oracle Change Management Pack, but it doesn't support global temporary tables in Oracle8.1.7, doesn't support the FLOAT datatype, etc., etc. What a disappointment that Oracle's own CM tool doesn't support their own database features. Go figure.

So, moving to the third party world, what appears to be the two best contenders at this point are Quest's Schema Manager (although it doesn't support global temporary tables either) and Embarcadero's DBArtisan Change Manager. I'm also using TOAD, EZSQL, and a freebie called SchemaCompare to perform a comparison of DEV and PROD after changes are applied.

Neither Schema Mangler nor DBArtisan Change Mangler appear to be perfect, although so far, DBArtisan Change Mangler appears to yield the best results. I've set up test schemas where I can apply the changes using the tools, and then compare the results. Of course, in the using the 3 compare tools I mentioned above, I get 3 different answers. I could just do all this with data dictionary queries, but with 1400 database objects, it becomes quite an over-whelming task.

Enough background, here are the questions:

1. Has anyone used or compared Quest Schema Mangler and/or Embarcadero DBArtisan Change Mangler? Is there another tool that I haven't found yet that works even better? Do you have any horror stories to relate regarding either tool?
2. Here's the most important question: If you compared change management tools, how did you go about comparing the source and post-change target after the changes were applied? Obviously you can't use the same tool (if it missed some detail the first time, it'll obviously miss the same detail the second time.)

Any info or opinions appreciated! Please post or email.

Regards,
Dave Received on Sun Apr 29 2001 - 20:38:25 CDT