There has been a lot of talk recently about needing to have a “modern architecture.” Many I have spoken with talk of needing an “analytic space” to either compliment or replace the “Data Warehouse.” Companies want to “embrace the cloud” but are not sure what that really means or what the trades off are that will occur in doing so.
The reason I put those in quotes is that they are vague terms open to interpretation. Unless the underlying discussions are held, companies can easily be led astray chasing after the next “silver bullet” solution.
The bottom line: Having a modern analytics architecture is going to mean something different for everyone. The key is understanding the capabilities you need to accelerate your business returns and then looking at the technology available today that can be leveraged to achieve those capabilities.
Capabilities of a modern architecture
Let’s start with what most people can agree on as to what a modern architecture should feature or enable. Your foundation needs to support:
- Raw to Connected to Integrated Data – throughout the data lifecycle
- Low friction / transparent access – even across different storage repositories
- Automated and dynamic management – with all dimensions of scale
- Flexible and portable deployment options to meet department or enterprise needs
- Resilient / seamless upgrades – minimized downtime
- Certified compliance to standards – able to leverage open source and first party tools
From the bigger perspective, you are really creating a system of systems and processes focused on accelerating business value. A complete ecosystem that all inter-relate and interact transparently from an end user perspective.
Modern is relative to time
What amuses me the most about the statement “modern architecture” is that modern is always relative to the current time. Twenty or thirty years ago, companies used the best of breed at the time and created analytic ecosystems, many of which they called a “data warehouse.” It was relational data in relational databases, able to accept SQL queries from a variety of front-end tools. These systems were running the latest CPU and storage systems and provided previously unknown performance and query capability.
Today, anyone thinking of building out their ecosystem would be laughed at if they were using the processors from the 1990’s or using “rotating rust” hard drives. While they were all the rage in the 90’s, they are not modern today. To complete the thought, whatever you build out today will be “unmodern” in 3-5 years.
It would be unreasonable to think that every 3-5 years you are going to completely re-do your analytic architecture from scratch, throwing away all you have in data flow, governance, optimizations, and management. Instead, leverage what is working and incorporate new technology where necessary.
Modernize, not Replace
The confusion today is that many new companies are trying to equate “being modern” with “total replacement.” It makes sense as they do not have the past successes and experiences, so the only message is that they are “new” and everything else is “old.”
But are you ready to throw out all the existing processes, optimization, and business analytics you have in place to move to a more modern deployment? I have heard that in a modern architecture you do not need to worry about ETL and optimization, just stream data in and let users run queries! That scenario has been tired and proven to fail. There are business reasons to integrate and optimize your data streams, just as there are reasons where having raw data is necessary. Going to modern should not mean abandoning all your current processes that are providing value to your business today.
Rather than a total replacement to be “modern,” it makes more sense to “modernize.” Look at your current deployment and processing systems. Are they leveraging the new cloud compute and storage options that allow for a tiered and performance driven architecture? Is your compute able to be accessed by newer environments of R and Python or do you need to extract data to these analytic silos? Do those compute resources span the gamut from simple reporting to sophisticated data science functions? All these capabilities are critical in a “modern” analytical architecture.
By modernizing your environment, updating where you can and adding where you must, you will carry forward the successful and business critical processes and governance that provide trust and transformation in the data, as well as achieve the move to modern in a much faster timeline with less risk.
Vantage – The modern Teradata offering
This brings me to the point that Teradata agrees that only having a SQL database, sitting in your data center running production reports on a fixed capacity system is not a “modern analytic solution.” Over time technology has evolved -- well so has Teradata, just as we have continually done over our 40 years of proven success and leadership in delivering analytic ecosystems.
Vantage is that next evolution of Teradata offerings, all the success of the past with the technology of today:
- New compute nodes so you not only have the relational SQL database but also machine learning and advanced analytics in a single environment.
- Separated compute and storage so you can independently adjust them as your needs demand, as well as incorporate other storage systems such as Hadoop or Object Storage.
- Platform independence so you can run the software in the cloud, on VM, on Teradata hardware, as a service, or any combination of these.
- Extension of API’s to allow a broader set of tools and languages which broaden the user community from business analyst to data scientist.
Vantage is the leading solution of choice in a flexible and portable world. It is not only a modern architecture today but sets up for the future which will be here soon enough.
And the best part is that you can achieve all this by upgrading to Vantage, and re-platform in the cloud. Teradata Vantage provides a simple path to being modern while at the same time bringing along all the innovation and capability that is driving your business success today.
For other perspectives on being modern, check this blog post by Chris Twogood.
Starting with Teradata in 1987, Rob Armstrong has contributed in virtually every aspect of the data warehouse and analytical processing arenas. Rob’s work in the computer industry has been dedicated to data-driven business improvement and more effective business decisions and execution. Roles have encompassed the design, justification, implementation and evolution of enterprise data warehouses.
In his current role, Rob continues the Teradata tradition of integrating data and enabling end-user access for true self-driven analysis and data-driven actions. Increasingly, he incorporates the world of non-traditional “big data” into the analytical process. He also has expanded the technology environment beyond the on-premises data center to include the world of public and private clouds to create a total analytic ecosystem.
Rob earned a B.A. degree in Management Science with an emphasis in mathematics and relational theory at the University of California, San Diego. He resides and works from San Diego.
View all posts by Rob Armstrong