Primary Use Case
As a contractor for various companies working in various industries, we usually find Teradata in financial organizations. It is often being used for relational data storage for huge data warehouses (usually built on Inmon principles) or as storage for building analytical data marts. Teradata is positioned as a high-performance solution for heavy analytical queries and, in fact, it is. Moreover, Teradata is typically supplied with the whole ecosystem of its tools (like Viewpoint), but I have not extensively used them.
Improvements to My Organization
- Cuts time to process huge amounts of data with analytical queries
- A conventional and easily defined way to build a data warehouse or a layer of data marts
- Better development time - Teradata can be easily used in ETL mode transformations, so there is no need for expensive and inconvenient ETL tools
Efficiency of running analytical queries. Teradata was mainly invented for that purpose and it really does it well. We were able to migrate the code from SAS analytical platforms to pure Teradata and got an order of magnitude of performance growth (taking into account that the hardware was similar).
Room for Improvement
Query language and its functionality are rather limited, compared to Oracle or even SQL Server. However, it is possible to perform any kind of logic in it (though some workarounds may be required).
Data ingestion is done via external utilities and not by the query language itself. It would be more convenient to have that functionality within its SQL dialect.
Use of Solution
Three to five years.
We have not experienced any stability issues.
No. Teradata is positioned as a rather scalable system, and we have not seen any issues with it.
Customer Service and Technical Support
I was not impressed by tech support. We made several requests regarding performance issues we had with our applications but did not get helpful answers.
It is hard to say what our clients had been using before. It is safe to say that this was their first option for analytical data warehouse implementation.
I cannot say for sure. We have not taken part in deployment activities. As it is an appliance, I expect it is a bit complex.
Pricing, Setup Cost and Licensing
Price is quite high, so if it is really possible to use other solutions (e.g. you do not have strict requirements for performance and huge data volumes), it might be better to look at alternatives from the RDBMS world.
Other Solutions Considered
Usual alternatives in the same segment for our clients are Oracle Exadata, SAP HANA, and Hadoop.
I rate it an eight out of 10. It is definitely the best RDBMS I have ever worked with, in the sense of building a data warehouse. It is fast, and easy to use in terms of writing queries. However, you have to understand the specifics of the data storage and its technical architecture, which is not typical for an RDBMS.
In terms of advice, I would suggest your provide training for your personnel, because it is rather important to understand the technical specifics of Teradata to create efficient solutions with it. Other RDBMS's usually requires knowledge of generic theory, but that will not work with Teradata.
Disclosure: My company has a business relationship with this vendor other than being a customer: Partner.