Each has its strengths and weaknesses; this is why we use a blended environment. We have a holistic view which is very mature and thus, it is impossible to give a concise answer.
However, converting the Teradata from a pure BI to an ODS has been our greatest achievement; it effectively has allowed us to remove over 20 portion copies of the data sets on other DB platforms for real-time operational reporting purposes.
Improvements to My Organization
It is the core of our fundamental real-time reporting and operational planning. It has saved us over eight figures in the operational run costs, improved data quality and consistency as well as reduced our overall infrastructure support requirements.
Room for Improvement
The increasing volumes of data demand more and more performance. I truly believe that in five years, we will breach the effective limits of Teradata to perform in a multi-petabyte environment, i.e., without significant changes to its underlying architecture. If this occurs, we will be forced to migrate to a more traditional platform (such as IBM Informix).
Use of Solution
We have aggressively been using BI for over 20 years, with this evolving environment. Our most recent setup was about three years ago.
We have a broad range of BI and reporting tools, built on top of the Teradata Platform. It, currently, is the active Data Warehouse on 12.x , although we are migrating to 13.x after the New Year. This is supported by our IBM Congos (12), MicroStrategy (10) and the QlikView (12) reporting tools. We have also built extensive analytics on a Java base as well as some Microsoft BI integration.
We have had stability issues quite frequently, but this is mostly with the Cognos environment, as it is growing beyond its limits in terms of the active user access. This is why we are considering migrating significant portions of our over 4000 users to a coded environment in place of the platform.
Again as mentioned before, there were scalability issues in regards to Cognos and the over 4000 users in a single active environment.
Customer Service and Technical Support
Our in-house team is world-class, so we are not reliant on any external support.
We have moved from a distributed heterogeneous to a centralized homogeneous environment. Costs, data quality, sharing of information, and supportability were the reasons as to why we switched over to Teradata.
The setup took place too long ago to remember.
Pricing, Setup Cost and Licensing
Negotiate, negotiate, and negotiate some more. But long before you start to have a clear understanding of your direction, use patterns and check the roles that BI plays in the organization. Make sure you have the in-house skills to design and support the solution, as relying on external sources is extremely costly and tends to lock you into specific platforms, tools, and paradigms.
Other Solutions Considered
We did a six-month review of potential candidates and options. Every three to five years, we re-evaluate the selection criteria and re-open the solution's design with an invitation to the vendors to offer an alternative. The process consumes about three months, but it has allowed us to keep current and, more importantly, to leverage off the process so as to ensure that we are getting the best value for our money.
This comes back to understanding the what, why, and how of the solution. Once you understand that, then plan it accordingly and ensure you have sufficient time to address any of the surprises that could occur.
Each tool fills a niche and is used accordingly. I doubt if there is anything perfect out there, as each tool has its own considerations when being deployed.
Our greatest concern with Teradata is the performance impact of the OUTER RIGHT JOINS across a petabyte environment.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Jun 22 2017