What is our primary use case?
Mainly, I have used Oracle Warehouse Builder in customer projects where the Oracle database environment is the main solution for data warehousing. It had been used mainly to collect and enhance data from other databases and pre-systems in order to establish data marts or similar data warehouse solutions for companies. The range of these pre-systems has varied from Microsoft-Office solutions to other databases (Db2, MS SQL Server, MySQL, etc.) and proprietary solutions (e.g., SAP).
How has it helped my organization?
The main intention besides having no extra costs for using Oracle Warehouse Builder has been the fast lane for developers getting access to (prepared) data sources and a midrange level of knowledge (Tcl, like scripting) to develop new ETL solutions supported by an easy to learn graphical user interface.
It becomes more challenging when diving into the details of data source management and more sophisticated logic. Therefore, expert or admin skills (PL/SQL, Tcl, and/or foreign data source specifics) have been required.
What is most valuable?
If the administrative setup was perfect, the use of Oracle Warehouse Builder with the graphical user interface including a wide range of supporting tool functionality could guarantee a fast and straightforward development and test process. In most cases, Oracle Warehouse Builder is flexible enough to fit for complex ETL processes and can be enhanced with extra functionality based on running a Java environment and database backbone structures.
OpenAPI supported me integrating Oracle Warehouse Builder with DevOps solutions like versioning (CVS, Subversion, and Git), ticket systems (Bugzilla and Jira), and test automatization (HPQC) as well as with professional ALM solutions (HP ALM and IKAN ALM).
What needs improvement?
According to the features above the Oracle Warehouse Builder was long time promised to be enhanced by a native versioning integration solution, but Oracle never got forward on this and preferred to force users switching to the paid solution Oracle Data Integrator, which urges a lot of work and challenges on switching from Oracle Warehouse Builder.
Unfortunately, a next release of Oracle Warehouse Builder will not happen but a wide range of customers still use the available features together with the current Oracle 11g or 12c database environments.
For how long have I used the solution?
What do I think about the stability of the solution?
As long as you don't change the environmental situation after a successful installation, then you will have a good DevOps concept. It runs for years with the normal support necessities in an Oracle/Java environment.
But we experienced some trouble every time we did major upgrades on the oracle database platform since common paths, environmental or database properties and some miscellaneous modules changed. So we had a lot of effort to fix this and even develop enhancements based on the Oracle API´s further because i.e. the kind of JAVA integration had been fundamental changed.
What do I think about the scalability of the solution?
Depending on the environment (database, service, Java, and VM), it is performant and scalable.
How are customer service and technical support?
Since Oracle provided a lot of still available documentation, we worked often together with Oracle Professional Services. I'm not able to guess how it would have been for "normal customers" in case of support. Now, no further support will be given by Oracle. However, there are still a lot of professionals worldwide which are able to support Oracle Warehouse Builder.
Which solution did I use previously and why did I switch?
I have used different solutions like DataStage, Informatica, and MetaSuite before and after. Besides MetaSuite, it gives me the full power of MetaDate management, code generation and platform independent use, optimized tailored integration for legacy systems (which was important for most of my projects), and open interfaces for individual DevOps solutions. Oracle Warehouse Builder was the best fitting option from a project worker's view.
How was the initial setup?
Since I could mange several standard installations on my virtual machines, it should not have been very complex. However, keeping in mind the dependencies to databases and Java environments, it should be setup together with administrative support in enterprise environments.
What about the implementation team?
The level of expertise should be an experience administrator or Oracle and Java user, at least.
What was our ROI?
Assume that customers would not have kept it for years if it would not have been profitable.
What's my experience with pricing, setup cost, and licensing?
Besides Pentaho, Oracle Warehouse Builder is a really low budget solution for those who own any kind of Oracle database license. From this point of view, it is often an unnecessary waste of time to evaluate other (pricey) solutions. Mostly, Oracle Warehouse Builder was the preferred solution because it did not cost extra for licenses when owning an Oracle database.
Which other solutions did I evaluate?
I have also tested Pentaho, Talend, and others but did not have any further project experience with them since I mostly had to follow customer strategies.
What other advice do I have?
If you have good experience with Tcl, PL/SQL, and Java, this could still be a valuable solution as long as Oracle technically supports the necessary database integration for the repository (up to 12c as far as I know). However, be aware that besides user groups and Google, you will not have any support for it in the future.
Which deployment model are you using for this solution?