We know from our science classes that gravity is the attraction between objects with mass. The bigger the mass, the more gravitational energy or attraction it produces. Just as the sun dominates our solar system because of its mass and energy, platforms within our data and analytic ecosystems can exhibit similar gravitational “pull.” Gravity is a fundamental element of nature and one that should also apply to data and analytics processing. In this sense, the combination of data gravity, along with the concept of analytic processing gravity
, come together to form a powerful force where users and applications are naturally drawn to a platform rather than having to be pushed.
Physical gravity is a naturally occurring phenomena and no matter how you try to avoid it, gravity will always be there keeping you grounded. With data, however, the fundamental laws of nature do not always apply. In today’s analytic ecosystems, the prevalent behavior is dispersion and replication where data can be spread far and wide causing less synergistic business activities and higher overall costs. A primary reason for this dispersion is that data often cannot be readily processed where it is stored or made accessible at the scale needed without having to yet again, replicate the data.
Even in cloud “big data” environments, where data gravity
can exist, data is often widely copied across the cloud-based infrastructure. What’s generally missing is the ability for analytics processing to act on data “in situ” or where it resides and is managed. In the cloud, data replication can occur even faster but not necessarily in an efficient way. This inefficient replication lowers the overall value of the data, analytics and platforms within a cloud-based ecosystem.
Leveraging data and analytic processing gravity is where Teradata Vantage has been proven to excel – be it on-premises or in the cloud. With many customers that leverage Vantage, more data, users and analytic development are attracted to the platform. And, not just because of ready access to data resources but also due to Vantage’s ability to process data efficiently while leveraging data gravity with performance at scale. What are the traits within Vantage that enable this attraction?
Mass and gravitational energy are closely aligned very much like data and analytic processing are related in our business universes. Analytic processing brings energy to data in order to produce valuable business results. Teradata Vantage has certain “performance at scale” characteristics and enablers that work to form data and analytics processing gravity. Some primary characteristics include:
- Ad-hoc queries and unfettered access to data – Any Query or Analytic, with Any Data for Any number of Users is readily supported by Teradata Vantage.
- Access to Any Data, virtually – Teradata Native Object Store and QueryGrid features support seamless access to data residing in other platforms. Vantage can drive additional data processing gravity by effectively bringing queries and analytic processing to data at scale – regardless of where it is primarily stored.
- Efficient processing and low aggregate processing costs – Vantage highly optimizes and orchestrates query and analytic processing to drive effective data processing. So much so that Vantage unit of work costs have proven to be much lower when compared to less efficient platforms.
- Workload throughput scaling – With other platforms when combined workload throughput hits the “gravity wall”, costly data and compute replication is the only solution. Vantage is proven to scale linearly and efficiently across single instances without inefficient and costly resource replication.
- Automation –Vantage automatically supports these characteristics without users or administrators having to be directly involved. Without increasing levels of automation, modern analytic process cannot occur given today’s growing requirements for data and analytic processing.
You may think that other platforms have similar characteristics, but these typically do not have the scalable data processing capabilities needed to generate gravity. Hadoop or object store-based Data Lakes are an example of platforms that can store “all the data.” However, the data often needs to be replicated, potentially many times across clusters or into other platforms to be processed effectively with analytics. Looking at the data level within and across these platforms, this division with hopes of conquering is proving to be a costly, failed approach. Without equivalent processing energy while limiting data replication, the value of data gravity is greatly reduced.
A true indicator of an effective and valuable analytics platform is one where users and applications are drawn to the platform and the platform can efficiently scale to meet and sustain this demand. This is enabled by the Teradata Vantage
characteristics listed above that create true data and analytic processing gravity where data use is supported at the scale needed without excessive replication. For Teradata Vantage customers where this type of gravity has taken hold, the value of the analytic data processing has proven to grow dramatically. And the overall benefit of this natural attraction should not be underestimated in today’s data and analytic-driven business world.
As a Senior Director & Principal Architect within Teradata, Bob supports a wide range of disciplines ranging from business consulting to in-depth technical services across many industries and business domains. He applies his 37+ years of experience and 25+ Teradata years with architecture and large-scale data & analytic implementations to help customers realize the full potential of their data and analytic investments. Bob focuses most of his current efforts on modern Data and Ecosystem architecture best practices.
View all posts by Bob Fintel