How to build a hybrid cloud by launching a more precise cloud data warehouse

By | 4 minute read | June 3, 2019

According to a recent IDC report, 79 percent of enterprises are currently investing in a hybrid cloud environment or have planned to invest in towards one in the next twelve months. More businesses are looking to do this by adopting public cloud deployments for their data management needs.

But adding public cloud to any previously on-premises data architecture can require months of planning and preparation as the first step. You have to answer questions like: What data can be stored on the cloud, and what data needs to be kept on-premises? What new data analysis tools or frameworks can you leverage, and how will these tools empower decision-making for the organization?

Making the business case for the agile enterprise

While a business works through the months of planning and decision-making to launch its hybrid cloud strategy, a common, first step to adopting a hybrid cloud is offloading data to a fully-managed, cloud data warehouse.

Companies often store massive amounts of data in on-premises warehouses aggregated from multiple sources— and with data of varying sensitivities. While businesses will need to keep their most sensitive data on-premises to comply with regulatory standards, there will be teams operating on a subset of this warehouse data that is suitable for the cloud. Unfortunately, as more users are added to the platform, the more contention there will be for on-premises resources—and likelihood of congestion.

In these cases, teams can move this subset of data to a cloud data warehouse to run in-depth analytics and machine learning on dedicated resources without disrupting core on-premises systems. Teams can get fast responses regardless of location. And the team utilizing the cloud data warehouse can leverage its elasticity to optimize performance and resource usage.

If the team needs to work on data volumes of less than a terabyte, then a smaller, more precise cloud data warehouse configuration should be used. We created IBM Flex One to help businesses do this by launching precise data warehouse configurations that utilize only the resources necessary to run high-performance analytics.

As an example, let‘s assume a major retailer is using the IBM Integrated Analytics System (IAS) to maintain sensitive data on premises. However, the data science team only uses demographic and general consumer data. With a few clicks, the retailer’s data science team can easily move a subset of cloud-suitable data to IBM Db2 Warehouse on Cloud using the IAS Lift to Cloud function.

Since IAS and Db2 Warehouse utilize the same common Db2 engine, as teams move data to the cloud, the team doesn’t have to make any changes to the queries they used to analyze the data when it lived on premises. They can seamlessly transition and burst to the cloud—without the need to change SQL dialects or adjust operational skillsets.

Making the business case for the growing enterprise

A smaller, more precise data warehouse configuration could also be advantageous for a mid-sized company experiencing immense growth. The current data warehouse market only offers solutions for a terabyte or higher. While growing businesses will need terabyte to petabyte size storage in the future, it probably doesn’t add up for their current business needs. The goal for growing companies should be to find a vendor who will accommodate their business needs and grow with them.

IBM created Flex One to help business operations entering a phase of growth to start small and grow storage and compute power as the organization expands its data. As an elastic data warehouse, Flex One allows you to scale your disk storage from 40 GB to 4 TB, and compute between 6 vCPUs to 28 vCPUs to accommodate varying workloads.

As businesses grow with Flex One into larger use cases and data volumes that require a more powerful data warehouse, they can request a Flex or Flex Performance configuration available on IBM Cloud and Amazon Web Service (AWS).

Offering Storage Compute
Flex One 40 GB – 4 TB (IBM Cloud) 6 – 28 vCPUs (IBM Cloud)
Flex 960 GB – 96 TB (IBM Cloud)

960 GB – 144 TB (AWS)

16 – 160 cores (IBM Cloud)

14 – 112 vCPUs (AWS)

Flex Performance 2.4 – 96 TB (IBM Cloud)

2.4 – 144 TB (AWS)

48 – 576 cores (IBM Cloud)

48 – 576 vCPUs (AWS)

Making the move to hybrid cloud

Moving to a hybrid cloud or fully-managed cloud can be a daunting task. But with the right configuration and growth plan, it can be simple and deployed rather quickly. The rewards for doing so far exceed this reduced level of hassle. For example, with IBM Flex One, organizations experience:

  • Embedded machine learning capabilities for deeper insights
  • Independent scaling of storage and compute to more exactly meet performance and cost needs
  • Self-service backup and restore to help protect against data loss
  • Kubernetes-driven architecture for enhanced high-availability
  • Common Db2 SQL engine for better integration with other data sources
  • Disaster recovery backups at no extra charge to help keep businesses up and running

All together, these capabilities form a highly-integrated, enterprise-ready, and forward-looking cloud option for small and medium businesses’ hybrid data management environment.

While IBM offers petabyte-scale data warehouse options for large enterprises running complex dashboards and reports, most companies kick off their data warehouse needs with less than a terabyte of data. If you’re at a small- or medium-sized business, you should consider using a smaller, more precise data warehouse configuration like Flex One.

Get started with Flex One within five minutes, or talk to your IBM Analytics sales representative to learn how Db2 Warehouse on Cloud can help your business grow.