Based on the ESG Research Survey conducted in partnership with Qumulo, Hybrid Cloud Trends and the Role of Active Unstructured Data, January 2020, 37% of organizations run AI in hybrid cloud infrastructure.
Leveraging AI in a Hybrid World
Cloud computing services like artificial intelligence (AI) and machine learning (ML) are incredible boons to research and higher education institutions, product development teams, or really any organization with a need to derive meaning from data at enormous scale. These serverless offerings allow users to get the benefit of the latest AI and ML capabilities without managing or even spinning up compute clusters. This lets users focus their efforts entirely on outcomes, not infrastructure.
Indeed, while cloud services are extraordinarily valuable, they are not easily accessible unless your data already exists within the cloud service you want to leverage. For some, this is not an issue, but for many organizations, the reality is that file data lives both on-prem and in the cloud.
Using a cloud-native enterprise file system like Qumulo allows users to span these environments, and locate file data where the service resides — even across cloud providers.
This is particularly important for users who rely on legacy software applications. It is challenging if not impossible to reliably move file data between legacy applications to native cloud services for processing in a manner that is practical and cost-effective for the business. For applications that are firmly rooted in the data center, this would require moving enormous volumes of file data to the cloud and back, while being mindful of the state of the data on-prem to ensure the two copies remain in sync.
With Qumulo, you can run the same workloads in the cloud as you can on-prem in your data center, because the Qumulo software running on-prem is the same Qumulo software that runs in AWS, Google Cloud, and soon Azure. This ensures feature parity in all environments, and the full enterprise file capabilities that many applications need to run properly. Windows and Linux workstation applications run without a hitch. Even the permissions you have set on files and directories work the same in the cloud, letting your Windows and Linux users collaborate on the same data without concern.
With the ability to seamlessly move your file data without compromise, a whole world of possibilities begins to emerge. Now, as you think about that, let’s add this fact: every single Qumulo software feature can be accessed via command line and our easy-to-use REST APIs, meaning that not only can you do it, you can automate it, too.
With the flick of switch (or a script), dynamically create a cloud instance in AWS or Google Cloud, include a Qumulo environment with the capacity and performance levels needed for the workload, and replicate your file data to it from your Qumulo file system on-prem.
Your legacy application file data is now in the cloud, running on Qumulo alongside cloud services. From there, share the data with any cloud-native service you like via NFS, SMB, FTP, or REST connection, and continuously replicate the results back to your on-prem Qumulo system as simply as it sounds. That is, simple.
Create and store your file data where it makes sense for your business without feeling boxed in. You don’t need to rewrite your applications to take advantage of cloud-native services—you need to rethink your file system. Qumulo is cloud-native enterprise file.
Download the ESG Research Insights Paper to learn more.
David Sniderman is a seasoned Technical Product Marketing Manager and former Product Manager at Qumulo with over 15 years of experience building enterprise, cloud, and mobile technology products. At Qumulo, David is focused on making our product benefits easy to understand and to use, and helping the tell Qumulo story in a big way!