Deployment without using enterprise etl options

Question

I've been burnt by Oracle again! After happily upgrading to 10.2 and shooting off "several toes" creating multiple control centers and configurations, I'm now ready to start deploying from dev to test to qa to prod.
While fighting several bugs in OWB regarding deployment, I see in the fine, fine print that multiple configurations are licensed as part of the enterprise etl option. Translation: another 100k plus to Oracle for something that I thought came out of the box for free. After all, we could deploy to various location using version 10.1.
So now I have to figure out how to move all these mappings and process flows we've created without a key component or pay Larry more money.
I need help! Other than using multiple configuration, control centers, and, I guess, a unified repository, what are my options?
Multiple OWB development repositories?
OMBPlus to hand write everything that the multiple configuration option would do? (I suppose that is a violation of the license agreement too)
Multiple control centers and tweaking the default_configuration every time we change something? (Again, I suppose I could do that with OMB)
SQL Script Exports and code hacks?
Any ideas or links would be appreciated!
-K

Answer 1

Hello.
Interesting problem eh?
I myself am going down this path, we have the Database licence so I am simply building a design repository in each region in my SDLC so one in DEV, TEST, QA, PROD. etc
I create a corresponding Runtime in each Region to match the Design Client. I do not use multi configuration in the Design Client!
For Release I migrate an MDL file and an XML deployment file. The MDL is simply for meta data publishing into next region, and if the needs arise it can be used for a hot fix and backward porting to DEV etc.
Using OMBPLUS i generate the MDL file and XML file in DEV region. Using OMBPLus I import MDL file into next region, and deploy XML file(DEV version) straight into Target runtime. Use OMBPLUS to sort out runtime connection details etc.
I Source Code control the XML and MDL files. So i have multiple layers of recovery points.
The concept is static XML file generated in DEV thats deployed into Each region up the chain. The DDL are being done in straight sql.
Using the Database wrapper proc I call the runtime and execute the mappings from dbms scheduler.
So the release process goes:
sqlplus DDL changes
OMBPLUS MDL import and XML deployment
sqlplus run schedules
So with consitent naming standards and development processes, its just the same release cycle with a bit in the middle to handle mdl and xml files via ombplus.
You get your transfer of responsibility from dev team to support team, dev team using ombplus to create the release plcode objects, the support team using the same ombplus programmes deploy the xml file into the target region and import the metadata.
You can mix and match how much, how little you want to use OWB.
Give us an update on how you proceed. As its an interesting topic.
Regards,
Richard.