How to Prepare Your Business for the Cloud-Optimized SAS Viya and Energize Your Analytics Program
03/18/2021 by Ivan Gomez Modernization - Analytics
Whether or not you use SAS as your data platform, the most recent release of SAS Viya has sent ripples through the analytics industry. Optimized for the cloud, Viya marks a comprehensive shift in how your business can draw insights from its data.
Easier to deploy across multiple servers while taking up less server space, Viya is the first cloud-native SAS product. Allowing for faster start-up times and increased security along with continuous integration and continuous deployment (CI/CD), Viya takes full advantage of the technology’s capabilities.
On the surface, the platform still offers the same familiar interface as its previous version. But from a functional standpoint, it’s as if a new model of a car was released with the same exterior but everything replaced under the hood.
Cloud optimized SAS Viya was released just last year and at the time was only available on the Azure platform. But in keeping with SAS’ traditional practices, SAS Viya will ultimately be platform agnostic as its rollout continues. If the expanded capabilities of Viya sound like they will provide a competitive edge for your business, you need to consider the steps toward a new purchase or a successful migration.
In its previous versions, SAS Viya was already a vital tool for delivering advanced analytics and compelling data visualizations. As a cloud-native analytics application, SAS Viya has streamlined deployment from developers to its end users. Viya’s adherence to the concept of CI/CD allows bug fixes to be resolved quickly with changes deployed as soon as you restart the software
But the speed and efficiency gains in Viya come down to its adoption of Kubernetes to manage its infrastructure. In the more traditional mode of maintaining analytics software, your data team would deploy all your scripts on a physical server and they would run there. SAS has redone the platform’s architecture with Viya.
Instead of relying on physical storage, Viya exploits the capabilities of the cloud. Kubernetes allows the platform to be containerized, which means the application runs in isolated spaces using the same operating system. An open-source container management system, Kubernetes functions like the captain of a boat deciding which of these containers go where.
Consequently, Viya requires less storage and runs that much faster than earlier Viya versions. The containerized architecture of this new version means less resources are required to stand up your environment and also makes it easier to manage.
With a smaller footprint and faster access times, Viya constitutes a big step forward for how your business gains insights from its data. But navigating the shifts that come along with the platform’s new architecture requires more than understanding the capabilities of the cloud.
Even if your business has already transitioned to a cloud provider for its flexibility, your prospective switch to Viya requires more specialized skills. Though Kubernetes powers the platform toward faster startup times and performance, it introduces very different concepts to your data infrastructure.
Much like any cloud architecture, Kubernetes operates without the need for a lot of additional overhead such as permanent storage or dedicated hardware. However in previous releases of SAS your organization would still need to reserve capacity in the form of virtual machines or dedicated instances that replicate the infrastructure of an on-prem environment. The Kubernetes-driven container capabilities of Viya enable it to be run at a lower cost because you can deploy it to a cluster that is highly scalable and extremely flexible.
Kubernetes containerized architecture means that storage is by default ephemeral. As a result, when a container is restarted or crashes all of the files on it are lost. That will not suffice for an enterprise-level application where users are creating code files, building models, and uploading data to the system. Persistent storage is key to a successful SAS Viya deployment. Therefore, it is important to use external storage solutions such as a Network File System (NFS) or Premium SSD Managed disks that will persist even when a Kubernetes cluster or container is removed.
However, even if you’re planning a migration from the latest version of SAS Viya the upgrade process can still be a challenging. Before taking advantage of the platform’s new capabilities, you need to do some homework. Becoming familiar with Kubernetes and its container storage companion Docker is essential before making the move. Learn the difference between a Pod and a Node and how to setup the networking that is required to enable access to your cluster from the outside. With this knowledge in place, managing your hosted environment and keeping up with maintenance becomes that much easier.
Once your team understands how Kubernetes works, the installation process for Viya is far simpler than for previous versions. Even if you’re a seasoned veteran with the platform, deploying the software alone could take days. During that time, the software could time out and lead to more delays. With software orchestrated by Kubernetes, Viya installations and upgrades will take less time and keep your admins happy.
For your users, SAS Viya looks just like the previous version, which eliminates the need for internal training. Really, only the structure behind the scenes of SAS Viya has changed significantly and most users will not notice the differences.
If you’re considering a migration from a SAS 9 release to SAS Viya, your migration can be more complicated. However, with the help of the right SAS consulting firm, you can make this upgrade to your analytics and storage capabilities seamless and efficient.
With the release of Viya, SAS has embraced the capabilities of a cloud-native application. But SAS has taken steps toward exploring the benefits of open source technology as well.
In the previous version of SAS Viya, the only part of the deployment that was automated was the initial installation. For Viya, SAS has introduced GitHub repositories for administrators to access. These repos provide an array of tools that can be used to automate infrastructure configuration, deployment, and the ensuing needs for monitoring and management of your software post-deployment. SAS administrators are sure to find these repos helpful as you create automated processes to assist in the configuration, deployment and management of Viya resources.
If your organization has completed a cloud migration as part of its digital transformation, the capabilities of SAS Viya are worth exploring. For larger organizations on the previous version of Viya with ready access to Kubernetes specialists, the migration is that much simpler.
But even if you’re a small- to mid-sized firm that’s already using Microsoft Azure, the barrier to entry is not impossible to surpass. Soon enough, the platform will extend to every cloud provider, which places a faster, cheaper means of managing your data analytics needs within your reach.