Rohit Amarnath is CTO of Vertica, a unified analytics platform that enables predictive business insights based on a scalable architecture.
Whether it’s scaling infrastructure, building software as a service applications, storing, processing and analysing big data, or managing backup and recovery, the public cloud computing industry has truly transformed businesses over the past decade or so.
Having said that, there are certain data management use cases that benefit from a hybrid cloud approach. In simple terms, the hybrid cloud data management approach leverages the power of both on-prem databases or enterprise data warehouses (EDW) and public cloud data warehouses (CDW). It often uses the very same technologies for both types of deployment so that transition from on-premises to public cloud is seamless.
Why Does Hybrid Cloud Data Management Make Sense?
Let’s face it, most businesses already have significant investments in on-premise hardware, and it often makes sense to leverage that data center in the most efficient way possible. What one chooses today (public cloud and private) may not always be the best solution tomorrow. By adopting a hybrid cloud approach, where on-premises looks and acts like the public cloud, businesses can enjoy greater flexibility and leverage the best of both worlds while making functional and financial sense.
Regulation plays a role in making this decision. When it comes to storing and processing data on the public cloud, some regulated industries require data and personally identifiable information (PII) to be stored on-premise. This is where a hybrid cloud makes more sense: Enterprises can choose to retain sensitive data on-premise with full security, control and authority while shifting other business-critical data and workloads to the public cloud.
Hybrid clouds are also a great way to avoid a single point of failure. In the case of mission-critical systems and applications, businesses can choose to run their regular workloads in the cloud while storing an on-premise copy for disaster recovery. Hybrid clouds also help avoid vendor lock-ins and expensive “egress fees”—exit costs when one tries to move their data out of CDW.
Seven Key Attributes Of Hybrid Cloud Data Management
According to Gartner, within three years, “85% of infrastructure strategies will integrate on-premises, colocation, cloud and edge delivery options, compared with 20% in 2020.” Hybrid cloud data management (HCD) will play a key role in identifying, unifying, harnessing and analysing data that will be stored across both on-premise and off-premise locations. Choosing the right architecture is also critical both for present circumstances and future growth, and that’s why IT teams must look out for these seven key attributes:
1. Integrated Analytics
All modern businesses wish to maximize the value of their data. Hybrid clouds must support analytical functions, tools and interfaces for business intelligence as well as advanced analytical techniques that help perform a variety of data exploration and visualization on applications, microservices, data pipelines and other components of the business.
For businesses to effectively transform customer experiences, modernize infrastructure and radically improve IT services, business stakeholders and decision-makers need a “single pane of glass” or a console from which they can manage different types of workloads and access previously siloed datasets.
3. Automated And Intelligent
By automating repetitive tasks, businesses can free up their data analysts and engineers for more innovative work. Hybrid cloud tech must therefore possess the ability to automate the configuration and management of its various components across the private cloud, public cloud and multicloud.
4. Fast And Scalable
The need for speed is driving consumer expectations and purchase decisions. Hybrid cloud architectures must enable short query response times (to meet rigorous SLAs), high throughputs (to query large volumes of data) and high concurrency (to support multiple workloads). As the business grows, so will the data, so it’s critical the technology is scalable enough to add more machines and data sources and to effectively distribute workloads across multiple clouds. However, scalability can’t simply be about adding more computing power when needed. It also has to offer tuning and workload management for the best value.
Hybrid clouds are built on the proposition of efficiency, and therefore, at their core, they should help the business conserve resources, especially cloud computing ones. For example, on-premise resources can operate steady, predictable workloads while public cloud resources can handle sudden spikes in demand (aka, cloud-bursting).
6. Open And Accessible
It’s pointless to collect data if the business cannot access, analyze or report on it. Organizations must ensure their data and workloads are accessible across an entire ecosystem of data stores, formats, processors, tools, libraries and APIs. IT teams should not have to write custom scripts to move data across these elements.
Legislative mandates like GDPR and CCPA are forcing organizations to raise their data and information security standards. The hybrid cloud must therefore provide cataloguing, auditing and tracking functionality along with securing data access with features such as role-based access controls.
There’s no debate that digitization and cloud computing are crucial to businesses that are trying to stay competitive and navigate future disruptions. The hybrid cloud data management approach is an excellent option for businesses that are vying for innovation and flexibility coupled with data protection, productivity gains and cost savings.