Archives of the TeradataForum
Message Posted: Mon, 21 Apr 2003 @ 17:12:09 GMT
One more thing to add regarding: Based on experience with other database systems I expect the system to start returning rows.
When querying DB2 the perception is that the response time is faster with a select * query due to the fact that DB2 returns pages of data where Teradata spools the entire results set before it sends it back to the user.
For example 2 million rows on each platform. A select * from each. When the user gets it's first response back from Teradata it is all 2 million rows being streamed back to the workstation where with DB2 it may only be a few hundred rows.
Response time should be measured by the over all time not the time it takes the first row to come back. That is why if you are going benchmark use complicated queries. This can give a better picture of your DBMS responsiveness.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|