Make the Jump to Cloud-based Data Warehousing with Hybrid Architecture
This spring, Teradata announced its hybrid cloud architecture capabilities. This architectural opportunity makes the leap to leveraging cloud-based data warehousing a smaller jump.
- By Jake Dolezal
- July 15, 2016
Attending the recent Teradata Influencers Summit, I saw firsthand their latest product offerings and road maps for cloud-based enterprise data warehouse (EDW) solutions. Although Teradata was certainly not the first cloud EDW out of the starting gate, their direction seems clear -- more cloud. According to their market research, by 2020 more than 90 percent of their customers plan to utilize the cloud for some portion of their data and analytics ecosystem.
Juxtaposed against those predictions is current reality. Business and IT leaders are trapped in indecision when it comes to analytics and how to meet their organization's data needs. Although the use-case for a cloud-based EDW is attractive, there are still barriers, and cloud proponents may have difficulty getting their organization off the fence.
On the one hand, the business needs to drive critical outcomes with agility and at scale while considering volume and cost. On the other hand, there is often significant pushback within the organization about security, architecture, and talent. Although many are not ready for a wholesale conversion to a cloud-based EDW, there are gentler and easier ways to start down the path.
When it comes to EDW architecture, the question is no longer "to cloud or not to cloud" because organizations can implement an EDW using a hybrid architecture.
That means the EDW can actually coexist on premises and in the cloud -- whether public, managed, or private cloud. In this case, the organization can use the cloud for load balancing by moving workloads off its on-premises EDW to an EDW in the cloud. With hybrid architecture, a company can also leverage a cloud EDW during periods of heavy analytics workloads.
Consider how during a month-end or year-end process the finance department historically brought the on-premises EDW to its knees. For this work (or all the other workloads), a cloud-based EDW instance could take the pressure off the on-premises EDW. Most likely, the users would be unaware this is occurring behind the scenes -- they would just enjoy their queries' good performance.
Teradata touts their current hybrid cloud capability; it uses cloud bursting to dynamically scale capacity among multiple EDW environments -- whether on premises or in the cloud. Bursting can dynamically move workloads from one instance to another. The dynamic workload movement can be automated by performance triggers or manually controlled by the organization.
The key ingredient to making this work is data synchronization. Currently, Teradata Hybrid Cloud allows for scheduled data movements to keep data in sync between the different EDW environments. Keep in mind, this does not mean all data needs to be moved and synced between on-premises and cloud EDW instances. More often, only the data needed to execute the dynamically reallocated workloads is synced.
A hybrid cloud architecture could let organizations enter into cloud-based data warehousing by leveraging the cloud to solve workload and performance issues. This is a much more manageable jump than making the difficult decision to completely migrate to the cloud.
Down the road, hybrid cloud capabilities could include being able to dynamically spin up a cloud EDW instance, migrate workloads, and sync small scale data on the fly. These capabilities make a powerful argument for considering a cloud EDW in the overall data ecosystem of your organization.
Dr. Jake Dolezal is practice leader of Analytics in Action at McKnight Consulting Group Global Services, where he is responsible for helping clients build programs around data and analytics. You can contact the author at firstname.lastname@example.org