Gartner has a vision for intelligent infrastructure. We complete it. READ THE PAPER >

The TidalScale Blog

    What is a Software-Defined Server?

    Posted by Gary Smerdon on Apr 23, 2019 7:45:23 AM
    Find me on:

    image (2)-1

    Ninety percent of all the word’s data did not exist two years ago. That’s how fast our collective data is growing.  So it’s fair to assume that sometime in the next 24 months, the volume of data your organization relies on will at least double in size.

    It’s also a safe bet that, if you’re not already concerned that the capabilities of your server infrastructure can’t keep up with data growth, you will be soon. The need to analyze all that data at increasingly faster speeds is growing right along with data volumes. In fact, achieving timely insights from big data has become essential to maintaining a competitive edge.

    This, in turn, has created enormous demand for in-memory computing solutions that overcome the performance limitations imposed by today’s enterprise storage infrastructure. The result: Many organizations are looking to scale their infrastructure to their data—without compromising flexibility and their ability to meet SLAs.

    The future is software-defined

    Whether the solution these companies seek is composable, converged, or hyper-converged, software-defined infrastructure is playing an ever-larger role in addressing the problem of circumventing the limitations of today’s rigid infrastructure. 

    Software-defined storage and networking have long been an important part of modern data centers.  But the heart of the data center, the server, has remained a fixed resource unable to be expanded beyond the limits of the individual computer.  To address increasingly memory-hungry workloads, some of the world’s largest technology companies and brightest engineers have been working for decades to find a way to make more memory directly available to the application to optimize performance.

    It wasn’t until TidalScale pioneered the Software-Defined Server that organizations were finally able to break through the traditional limitations of in-memory computing. A Software-Defined Server gives you the ability to combine the resources (memory, processors, and I/O) of multiple existing servers into a single virtual server – under a single unmodified operating system image.

    Need help meeting the challenges of Big Data?  Find it in our new paper featuring research from GARTNER >

    5 characteristics that are core to Software-Defined Servers

    To deliver real value, a Software-Defined Server must meet certain criteria—anything less is a compromise that will cost organizations time, money or both. A Software-Defined Server must:

    1. Deliver a server of any size, completely on demand. The solution must provide you with the ability to flexibly configure and deploy servers of any size - from tens of gigabytes to tens of terabytes of system memory. If you’re looking at a software-defined solution that places limits on the configurations of your virtual servers, keep looking.

    2. Rapidly reconfigure to adapt to evolving user needs. When user needs shift, you need to tear down your current virtual server and configure a new one just as quickly— in minutes. If you’re looking at a solution that could threaten SLAs by preventing you from quickly reconfiguring your Software-Defined Server, keep looking.

    3. Compose a system using only standard commodity servers connected with Ethernet.  Data centers achieve peak operational efficiency through delivering homogenous infrastructure, at scale. This is also the key to enabling a flexible Cloud/on-premises IT strategy. Complex, rigid proprietary infrastructure is inconsistent with the goals of Software-Defined Servers. If you’re looking at a software-defined solution that isn’t built to run on industry-standard hardware and software, keep looking.

    4. Run standard software. Software-Defined Servers need to run standard software—unmodified applications on unmodified operating systems—and share the same vendor certifications as traditional servers. They also must provide a programming model that eliminates the need for complex scale-out applications. If you’re looking at a software-defined solution that introduces unnecessary complexity or forces you to modify your code, keep looking

    5. Operate in the Public Cloud or on premises. Flexibility, cost and ease of use make software-defined infrastructure valuable to an IT organization. That should apply not just to IT assets you have on premises, but also those you rely on in the cloud. If you’re looking at a software-defined solution that disrupts your cloud strategy, keep looking. 

     These are all critical for IT organizations to realize the value of Software-Defined Servers and complete their vision of a truly software-defined data center.  And they are the core guiding principles that TidalScale used when pioneering Software-Defined Servers. Find out more by reading our latest whitepaper, The Next Wave: Implementing Intelligent Infrastructure with Software-Defined Servers, featuring research from Gartner.

    Or contact us for a customized demo of TidalScale software.