In my previous post titled Top Five Reasons to Get Started With Open Networking and SDN, I ended with Reason #1, “Get more out of your networks,” including more performance and scalability, more flexibility and choice and more operational efficiency. Today let’s explore how you can get more of all those things.
#1. Plan a Smooth Migration
How do you start using open, software-defined networking (SDN) without disrupting existing network operations? Many SDN approaches don’t have a good answer to that question – they’re OK for “greenfield” deployments, but if you want to upgrade your existing network, they force you to “rip and replace.”
There is a better way. Smarter SDN solutions, such as Pluribus Netvisor® ONE and Adaptive Cloud Fabric™, allow you to start small, with just a few top-of-rack switches or even a single switch that interoperates seamlessly with your existing network (Figure 1).
Then when you’re ready, you can extend your new SDN fabric across your entire network (Figure 2). And you can do it without touching your existing spine switches and wide area network (WAN).
#2. Rethink the SDN Controller
Most people think “SDN” means an external controller, or more realistically, several controllers, to achieve acceptable levels of performance and availability. That means complexity and cost, and for most networks, it’s not worth it.
Once again, there is a better way: controllerless SDN. With the Pluribus Adaptive Cloud Fabric, the SDN control plane is fully distributed across every switch in the fabric, so there is no need for an external controller, and no need for special proprietary protocols between controllers and switches. This approach has profound advantages:
- Avoid the cost and complexity of multiple external controllers
- Eliminate single points of failure and increase resilience
- Avoid closed, inflexible architectures and proprietary protocols
#3. Simplify Everything with a Unified Fabric
Open networking without SDN can reduce hardware costs, but it doesn’t necessarily help simplify network operations or lower operational costs. The network still must be configured box by box. New services must be provisioned box by box. Troubleshooting is box by box. You get the picture.
There is definitely a better way. When you log in to the Pluribus Adaptive Cloud Fabric from any switch, anywhere in the network, you see the entire network (Figure 4). With fabric-level service objects, you can provision any service or any policy across the entire fabric with one touch, eliminating as much as 95% of operational cost and dramatically reducing the potential for manual configuration errors. With fabric-wide visibility, telemetry and analytics, monitoring and troubleshooting are greatly simplified. Perhaps more importantly, fabric-wide automation enables you to move faster and be more agile in responding to your users and customers.
#4. Embrace Segmentation and Slicing
There’s a lot going on in your data centers. To support all your different applications and users with full security and privacy, you often need a way to isolate them and give them their own slices of the network that can’t disrupt other users and applications.
Full isolation between customers is clearly required in the multi-tenant data centers of managed service providers (MSPs) and cloud service providers (CSPs) such as Econocom. The Pluribus Adaptive Cloud Fabric enables deep slicing – down to the hardware layer and across the entire fabric – to create isolation across control, data and management planes (Figure 4). This is not only critical for multi-tenant security; it can also enable much more individual user control over each network slice.
It’s also highly valuable in other use cases, including rapidly expanding applications for the Internet of Things (IoT). Steelcase puts untrusted IoT devices and cloud-based IoT applications in their own network slices, isolated from the enterprise network (Figure 5). This reduces security risks and enables better application performance management.
#5. Automate, Automate, Automate
What’s better than simplifying operations? Automating them! Network operators trying to keep up with dynamic applications and unpredictable traffic demands increasingly need network automation tools to help.
The good news is that the Adaptive Cloud Fabric is ready for automation. Everything you need to do in the Adaptive Cloud Fabric is supported via a simple, open REST API, which enables integration with any automation or orchestration framework. Integrating with VMware or OpenStack? We’ve been there and done that. Want to use your Ansible tool kit? We have playbooks for you. What’s more, the simplicity of fabric-wide “one-touch” operations extends directly to simpler integration, letting you automate an entire network as simply as a single switch.
#6. Increase Your Insight with Analytics
No matter how much you succeed in automating your network for hands-off operations, there are times when you need to get hands-on to figure out what’s going on. Whether you’re trying to improve application performance, troubleshoot a temporary outage report or investigate a suspected security breach, you need tools that help you figure it out fast.
Pluribus Insight Analytics delivers powerful visualization and analysis capabilities built on the robust traffic flow data streaming from the switches in the Adaptive Cloud Fabric. With a unique capability to capture and analyze every TCP flow, to look back hours, days or weeks – all without the prohibitive cost of additional hardware – you can get the insight you need, quickly, to keep your network operating smoothly and securely.
As you can see, Pluribus helps customers get more from their SDN-based open networking transformation. Can we deliver more for you? Contact us.
Want to learn more about open networking and SDN?
Subscribe to our updates and be the first to hear about the latest blog posts, product announcements, thought leadership and other news and information from Pluribus Networks.
About the Author
Jay Gill is Senior Director of Marketing at Pluribus Networks, responsible for product marketing and open networking thought leadership. Prior to Pluribus, he guided product marketing for optical networking at Infinera, and held a variety of positions at Cisco focused on growing the company’s service provider business. Earlier in his career, Jay worked in engineering and product development at several service providers including both incumbents and startups. Jay holds a BSEE and MSEE from Stanford and an MBA from UCLA Anderson.