Hybrid clouds seem to be the way the wind is blowing. According to the Enterprise Strategy Group, the number of organizations committed to or interested in a hybrid cloud strategy has increased from 81% in 2017 to 93% in 2020. But what exactly is a hybrid cloud? Turns out, there are a lot of definitions. I’ll share a definition from Deloitte that I like:
“Hybrid cloud is cloud your way. It’s integrating information systems—from on-premises core systems to private cloud, public cloud, and edge environments—to maximize your IT capabilities and achieve better business outcomes. It’s designing, building, and accelerating your possible.”
Why Hybrid Cloud?
Data warehouse deployments on-premises and in the public cloud play equally important roles in a hybrid cloud strategy. The Enterprise Strategy Group found that 89% of organizations still expect to have a meaningful on-premises footprint in three years. At the same time, Gartner predicts that public cloud services will be essential for 90% of data and analytics innovation by 2022. Accordingly, organizations are adopting a hybrid cloud strategy to leverage the right mix of locations to meet their needs.
Consider: The cloud provides the flexibility to build out and modify services in an agile manner, the potential to scale almost infinitely, the assurance of enhanced business continuity, and the ability to avoid capital expenditures (CapEx)—all of which continue to accelerate the adoption of cloud-based data warehouses. But data warehouses running on-premises in your own data center deliver their advantages:
- Data gravity: Sometimes data is hard to move to public clouds since there’s so much of it and/or the data have interdependencies with other systems and databases.
- More control over governance and regulatory compliance: You know where and under what geographic or other restrictions your data is operating.
- More control over deployment infrastructure: You may want to use hardware, operating systems, databases, applications, and tools you already have.
- Avoiding high operational expenditure (OpEx): Consumption-based pricing models in public clouds can lead to high OpEx when usage is frequent – particularly if that data is fluid, moving between public clouds and on-premises locations.
Hybrid Cloud Evaluation Criteria
To get optimal benefits from a hybrid cloud data warehouse, though, you’ll need a solution that can drive better business outcomes while using fewer resources. For starters, you’ll want a single-solution architecture that can operate in both public and on-premises environments. Solutions from many data warehouse vendors either don’t do this well or don’t do this at all. Many vendor’s data warehouse solution runs in the public cloud or on-premises, and their “hybrid” versions have been cobbled together to meet the increase in demand. However, without the same data and integration services on-premises and in the cloud, the same data model, the same identity, and the same security and management systems, these solutions effectively saddle you with two siloed data warehouse deployments.
Why are common data and integration services, the same data model, the same identity, and the same security and management systems important? Let me tell you:
Same Data Services
It is essential that your data warehouse supports the same data services for public cloud and on-premises data warehouses. Without this, you will wind up with data redundancy and data consistency issues, duplications of effort and resources (human and technical), increased costs, and an inability to provide seamless access across environments.
Same Data Model
A data model determines how data is organized, stored, processed, and presented. Having a single data model for the on-premises and cloud-based sides of your data warehouse eliminates incongruencies across source systems. It also strengthens data governance by ensuring that data is created and maintained in accordance with company standards, policies, and business rules. As data is transformed within the data warehouse—on-premises or in the cloud—it continues to adhere to data definitions and integrity constraints defined in the data model.
Same Identity Authentication
Your users should be able to sign on to on-premises and cloud data warehouses using the same login ID and password. Data warehouse support for single sign-on (SSO) access helps eliminate password fatigue for users and, more importantly, can help you ensure that your organization’s identity policies are extended to protect both data warehouse locations.
Same Security and Management Services
Shared security services are also critical for hybrid cloud data warehouses. I’ve already written two blog posts that provide details on security, governance, and privacy requirements for the modern data warehouse, one on database and one on cloud service security, so I will refer you to those for more details. But I would like to point out in this discussion that you will need integrated security services across your on-premises and public cloud environments to ensure a strong and consistent security posture for your hybrid data warehouse.
Finally, shared services for management tasks offer clear advantages in terms of cost, control, and simplicity:
- You’ll need fewer staff members to develop, maintain, and monitor the components in your hybrid deployment.
- You’ll improve control through consistent upgrades, patches, and backups.
- You’ll simplify metering and licensing requirements across environments.
Actian Avalanche Hybrid Cloud Data Warehouse
It should come as no surprise that a hybrid cloud data warehouse that meets all these criteria does exist: the Actian Avalanche™ hybrid-cloud data warehouse, integration, and management platform is a true hybrid cloud data warehouse that can be deployed on-premises as well as in multiple public clouds, including AWS, Azure, and Google Cloud. You can read more about the Avalanche solution here.