Oracle has added Sunopsis to their Fusion middlewear line of products. Congratulations I think. On the surface this looks good and adds credibility to the already increasing evolution of ELT from ETL. The question will be whether this is too early in the adoption cycle.
What I am thinking is that would Teradata and Netezza want to promote this product (whether it is good or bad) over other ETL tools when they would be introducing a foothold for a competitive product. Hummmm.
Ultimately the consumer should win if they can get over that possible issue. Take IBM and DataStage. It certainly supports the consolidation effect across the Enterprise software space. I think Oracle hit a homerun on this product.
Update: Vincent has a very good post and some good converation started over here
I know this is an old thread, but that’s the beauty of blog search I guess.
About the idea of Teradata and/or Netezza introducing the Sunopsis solution – Oracle is going to be there anyway. If they’re not directly in the competition, they’re somewhere else in the organization. Besides, if a customer is using Teradata already, the Sunopsis solution can be a good way to improve their incumbentcy (sp?). You have to believe that at least at some level, doing the right thing for your client pays dividends.
While I have to admit Stephen has been trying to get me to look at ELT tools like Sunopsis for some time now, I don’t see the need for them. If I’m going to do my heavy lifting inside the database, which is certainly the proper place to do it in my opinion – especially if you’re using the right horse power engine i.e. teradata. Then I can do most of that heavy lifting with SQL and good set manipulation skills in macros, triggers, and procedures that come imbedded in my database. And if I’m looking for metadata from all that well I have a challenge there no matter what tool I use because no vendor has cracked that metadata nut to my satisfaction as yet.
Larry (Ellison) has another tool in his chest of toys, right beside the MIG (Russian fighter jet the US didn’t let him buy) and the Americas Cup(- oh yeah he didn’t get that either) 🙂
Some good conversation.
I was refering to the challenge of Sunopsis no longer being an independent tool and that it would be hard in a sales situation for someone to want to introduce Oracle if they did not need to. The challenge is that if Teradata or Netezza brought this tool in now, it would immediately bring in Oracle with legitacy in the overall solution allowing for a quick sale of an enterprise license and the appearance of a cheap path which might diverge from the architectural strategy.
So if I was a Netezza or Teradata rep, I would have a hard time wanting to introduce this into my solution structure, assuming I have options.
As mentioned I am a big fan of Sunopsis and like Oracle, but found that the maximum bang for the buck was Sunopsis on Teradata as they have such good built in transformation tools that Sunopsis leverages very well.
Good point Jim. OWB is Oracle Warehouse Builder, the Extract-Load-Transform tool Oracle built internally. Looks like it will get merged with the Sunopsis ELT tool. It will have to be renamed so it doesn’t sound like a tool that only loads Oracle data warehouses.
I know you guys are talking to each other but could you, please, at least give the meaning of the three letter words the first time you use them. The aim of these little items is to talk to the uninformed rather than the people who already know this stuff. Help!
I think ETL is extract, transform, load. ELT is extract, load, transform. But have no clue what OWB is.
I always wonder when one software company buys another, whether the two products really can work together. Often look good on a features page, but really do not fit when they are put together without alot of work.
According to the Oracle Q&A they are going to keep the Teradata partnership with Sunopsis going (see my recent blog post for details).
My guess is that Teradata have a lot to gain from ELT implementations. An ETL approach puts a lot of CPU load, staged data and intermediate tables on the transformation servers resulting in a lot of money spent on ETL CPU based software licenses. The ELT approach sees the load and disk space put on the Teradata server resulting in bigger Teradata installs and more revenue. They have built a very rich layer of transformation functions which helps the ELT scenario.
If Teradata lose the partnership with Sunopsis, for example if Oracle push the ELT into OWB, they can fall back on the slightly less lucrative partnerships with the ETL vendors. Many of these vendors have some ELT support anyway and a lot of sites are doing the data warehouse aggregation builds within Teradata and not on the ETL server.