Virtualization Technology News and Information
SnapRoute 2019 Predictions: Time to Work Together Towards an Operator-First, Cloud-Native NOS

Industry executives and experts share their predictions for 2019.  Read them in this 11th annual series exclusive.

Contributed by Adam Casella, co-founder SnapRoute

2019: Time to Work Together Towards an Operator-First, Cloud-Native NOS

Network Operations is not known as the most enticing of fields. The behind-the scenes work is complex and mission-critical, and doesn't involve creating "lit" content for social media. But the rep of NetOps is due for a refresh in 2019, as network infrastructure innovations catch up with cloud tech.

One of the many reasons that networking is seen as an endlessly frustrating endeavor is that the industry has been vendor-centric for far too long. Because network operating system (NOS) architectures have traditionally been monolithic, update and innovation cycles are far too long. Major maintenance requires downtime, which everyone hates, especially on the business side. And NOS complexity means the network ends up being a roadblock - it can't readily delivery on the promises of cutting edge application infrastructure, and it wasn't designed to handle automation.

In 2019, it's high time to figure out what an operator-first NOS looks like. It's imperative to create a NOS approach that aligns with virtualized compute, cloud services, and DevOps in terms of agility, scalability, customization, and security. Moreover, we need to solve problems around reliability, performance, and cost. The pressure on network and app teams to deliver better time-to-service is enormous.

Applications are demanding, more portable, and more dynamic than ever before - think of what you can do with a SaaS solution that would have been unimaginable 20 years ago - but the network hasn't kept up. Networks simply aren't responsive to the changing needs of applications. We've been working on the same issues for a decade, but progress is (too) incremental. We need to solve for Day 2 Operability.

As I see it, the near future of network operations will be characterized by tackling central challenges with innovative, modern approaches like containers, microservices, automation, and machine learning. The primary challenges include:

  • Network and application complexity
  • Network and performance visibility
  • Scalability with stability
  • Consistency of deployment
  • Convergence time
  • Runtime serviceability
  • Security and compliance
  • Cross-team collaboration
  • Human error from manual methods

In other words, 2019 is the year to apply Cloud Native principles to the NOS. This helps bring the management of development, operations, and network into alignment, which will cure a host of ills. If we can get these teams on the same page, we can start sharing knowledge and ideas, working efficiently towards common objectives, and fostering an environment where agility is rewarded, and innovative measures aren't dismissed as impossible risks.

Some attempts have been made to drag networks into the cloud era, but they fall short. First, trying to represent network models in a server-like way in order to apply DevOps principles fails to acknowledge that network devices have unique characteristics and needs. Second, disaggregation models still boil down to monolithic management: the same traditional NOS authors have not addressed the main issues, and the security model has not changed.

We can do better! We've seen it in action with the rise and success of DevOps and other Cloud Native approaches. Operators want to innovate and add features at their own speed, based on their particular needs. By applying Cloud Native principles to the NOS, we can utilize containerization and microservices to intelligently segment protocols, functions, and infrastructure on a network device, whether it is on premise, or in the cloud. We can empower operators with feature independence and employ true CI/CD to drive configuration changes and other updates needed to optimize network performance, availability, and security.

This is no small task. It requires network operators to work together to create approaches that allow for innovation and flexibility from day one. We can't simply push a single solution (no going back to vendor-first paradigms!) or modularize a monolithic solution and call it good. The industry needs a platform designed from the roots up on a microservices architecture driven by application workloads.

Here's where we hope and expect to see progress in 2019:

  • Leveraging of the large open source community around microservices to create innovative components for the network
  • Make the network operationally transparent and native to the DevOps toolchain
  • Enable high value, distributed services to be delivered natively in the network
  • Deliver on the promise of application infrastructure automation by clearing network roadblock
  • Increase agility, scalability, reliability and performance while decreasing cost - hand the steering wheel back to the operator

A Cloud Native NOS will modernize the networking industry, foster better alignment with development and applications teams, attract fresh talent and ideas, and prepare the network for the next major step change in technology, whatever that may be. It will allow for a common operating model between DevOps and NetOps and bring the power of CI/CD to network configuration and security. There's no need to change the underlying protocols - we're don't need to fix what isn't broken. And we aren't starting from scratch - we have cloud technology successes and DevOps models to guide the way. We need to catch up to compute and put operators first. We're looking forward to working across the industry to make great strides in the year ahead.


About the Author


Adam Casella co-founded SnapRoute in August, 2015, where he is responsible for setting technical direction and ensuring the operator's perspective is always seen. Adam's background of supporting networking products as a vendor, paired with his operational experience running large-scale data center networks gives him a unique perspective on how to build reliable, resilient, and easy-to-use products. Adam is an authority on both disaggregated and traditional networking technologies and especially their use in hyperscale spine/leaf CLOS designs and topologies.

Prior to founding SnapRoute, Adam was responsible for designing and building hyperscale data center networks at Apple. This included the full spectrum of operations from product evaluation, proof-of-concept labs, overall topology design, configuration schemes, device deployment, and maintenance. Before Apple, Adam was a lead engineer in Cisco's TAC on the LAN and Data Center Switching teams.

Adam has a BS in Network and Systems Administration from RIT (Rochester Institute of Technology).

Published Monday, October 15, 2018 7:57 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<October 2018>