Difference between ORABULK stage and RBDLOAD stage

dprasanth · Post by **dprasanth** » Tue Oct 18, 2005 6:40 am

Hi,
I was trying to bulk load a file with 5 million records.First thought I will do it by unix command line sqlldr command. Then since datastage has that facitlity I thought I will use that. I went through the help files which mentions to use ORABULK stage. But what we have is RED BRICK stage. Is there any difference between them or both are one and the same.
Thanks in Advance

chulett · Post by **chulett** » Tue Oct 18, 2005 7:00 am

No, they are not one and the same - they perform similar functions, but for two totally different databases. So, if your target is Oracle you could use the ORABULK stage. Similarly, if your target is Red Brick you could use the RDBLOAD stage.

Stick with command line sqlldr. Just because you've got a stage that does something doesn't mean it's the best way to do that something.

dprasanth · Post by **dprasanth** » Tue Oct 18, 2005 7:01 am

chulett wrote:No, they are not one and the same - they perform similar functions, but for two totally different databases. So, if your target is Oracle you could use the ORABULK stage. Similarly, if your target is Red Brick you could use the RDBLOAD stage.

Stick with command line sqlldr. Just because you've got a stage that does something doesn't mean it's the best way to do that something.

Thanks for the clarification

dprasanth · Post by **dprasanth** » Tue Oct 18, 2005 7:12 am

chulett wrote:No, they are not one and the same - they perform similar functions, but for two totally different databases. So, if your target is Oracle you could use the ORABULK stage. Similarly, if your target is Red Brick you could use the RDBLOAD stage.

Stick with command line sqlldr. Just because you've got a stage that does something doesn't mean it's the best way to do that something.

Actually I didn't know that there was a database called Redbrick. That is why I had a doubt between those two stages. Just now found out that Redbrick is a database from IBM for business Intelligence application. Thank you so much for the guidance

chulett · Post by **chulett** » Tue Oct 18, 2005 7:13 am

No problem. You'll find there are alot of IBM peoples who don't know there is such a thing as Redbrick, too!

kcbland · Post by **kcbland** » Tue Oct 18, 2005 7:24 am

IBM only owns Redbrick because they acquired Informix, who only owned Redbrick, because they acquired it from Ralph Kimball.

ray.wurlod · Post by **ray.wurlod** » Tue Oct 18, 2005 4:25 pm

Red Brick should be THE database of choice for star schema designs (data warehouses and data marts) because it was designed from the ground up to do exactly that. By Ralph Kimball, acknowledged as one of the masters of data warehousing theory and practice.

Red Brick is not a transactional database. One of its performance secrets is the amount of stuff it doesn't have in it. Another is the ability to create an index over multiple tables, so that joins can be resolved in the index; by this means sub-second query response times can be had from constrained queries on billion row fact tables joined with several dimension tables. Its bulk loader can use parallel processes to update indexes as the main table's rows are being updated, and still other parallel processes to update aggregate tables automatically.

Another brilliant feature is that it monitors queries that do and could use aggregates, and can recommend that particular aggregate tables ought to be in place. Users don't need to see them; Red Brick automatically rewrites queries to use aggregate tables if appropriate aggregate tables exist.

Finally, because of what it is, Red Brick applies "query priority concurrency". Queries always get precedence over updates. And, if one process is scanning an index or table and another process needs to, the second process will "piggy back" the first, minimizing the amount of work that needs to be done. And there are lots of useful business intelligence functions in its SQL, such as ranking, tiling, moving sum and moving average, and so on.

But, alas, the news hasn't filtered through to the IBM sales dudes, who've been brainwashed into believing that DB2 is the only database the world will ever need. Or is it that the margins on Red Brick aren't nearly so large?

Here endeth the rant.

chulett · Post by **chulett** » Tue Oct 18, 2005 5:55 pm

We feel your pain, Ray.

kcbland · Post by **kcbland** » Tue Oct 18, 2005 7:56 pm

Silly little dreamer.

Who says the best product has to win?

Without a doubt, properly modeled and properly adminstrated, Redbrick is phenomenal.