On 10/13/06, Ric Van Dyke <ric.van.dyke@xxxxxxxxxx> wrote:
I worked with a couple of Teradata databases many years ago. Back then Teradata was better at handling huge data volumes. Today they don't have the advantage they used to. The Teradata is built on a "shared nothing" architecture. The system is made of a set of things called an AMP. This AMP is basically a computer system of its own, and it's kinda like clustering. However the AMP don't share any resources. Each AMP will be the master of a part of the data. No one AMP has access to all the data. The data is hashed across the amps and this hashing is the key to Teradata. Pick a bad hash key and life really isn't that beautiful. Pick a good hash key and that data is well populated across the AMPs and life is good.
I believe it's still a proprietary system it has to run on their hardware. I know they did reconfigure the hardware such that other databases (including Oracle) could run on their machines.
Teradata's implementation of SQL wasn't very robust back then, maybe its better today. But Oracle really had Teradata beat in that world for sure. And the tools available were very "green bar" paper orientated. Its been 10+ years since I worked with Teradata so much of this has likely changed.
Teradata was build for and is still a good database for huge databases. Oracle start with the smaller databases and came up to large databases, and for a while wasn't a player in that area. But those days are gone and the folks who still talk that Teradata is the "only database" that can handle huge data volumes are stuck in those days.
Ric Van Dyke
Hotsos Enterprises
-----------------------
Hotsos Symposium March 4-8, 2007. Be there. ------------------------------
*From:* oracle-l-bounce@xxxxxxxxxxxxx [mailto: oracle-l-bounce@xxxxxxxxxxxxx] *On Behalf Of *cosmin ioan *Sent:* Thursday, October 12, 2006 6:57 PM *To:* oracle-l@xxxxxxxxxxxxx *Subject:* re: Teradata Comparison
hello all,
for the past few weeks, I have not heard the "end of it" in a number of discussions I've had with a developer holding in high regards Teradata as being the "serious" datawarehousing db, "much better" than Oracle. Not knowing the particulars of Teradata, naturally, I could not digress or make a comparative point, however, to me, even the biggest, the uber-googol databases, can be partitioned & summarized pretty efficiently for awesome results & response.
Has anyone worked with Teradata? What are some of their strenghts that Oracle does not provide? -- I doubt there are many, if any at all, but I'm looking for a more expert opinion.
thx a bunch,
Cosmin
