What do you like best?
Wow! There are so many things.
-> I guess I can start with MPP architecture. Yes, of course other database shave caught up to Teradata with that regard.
-> easy syntax. The syntax of Teradata is clean and pure. It is not as complicated as SQL Server or Oracle.
-> Used extensively for Datawarehousing and Business Intelligence Applications
-> Primary and Secondary Index for Data Storage and Data Retreival
-> Full table scanning is a real possibility in this table, without taking much time.
-> Fastload, Multiload capablities couples with t-pump for near real time operations.
What do you dislike?
Having worked on this database for quite a few years, there is very little that I do not like about this product.
-> The fast load and multiload are currently a set of scripts. If this can be somehow made into a GUI that can enable easier load of data (like the fastload and multiload supported expressions in other ETL tools) it could be useful.
-> May be if at all there was anything to be worked on (like 1/10 chances if any changes would be made), may be the GUI (the visuals) could be made a little more appealing.
Recommendations to others considering the product
Great Product. Depends on the application to be used. If the company needs massive data to be queries and reported on then go for Teradata. If the data volume is not too high this could be a costly expense.
What business problems are you solving with the product? What benefits have you realized?
I have worked on more than a couple of projects dealing with Teradata Database. The best of the implementations was as a Data Driven Decision making application.
-> Can load multiple sources of data
-> Unlike SQL Server or Oracle, there was no need to write PL/SQL. SQL DML was sufficient to suffice the requirement.