I have been in networking my entire career, first as an engineer, then as a business executive, and now as CEO of Apstra. In 2014, I started Apstra with Sasha Ratkovic, a thought leader in domain abstractions, and David Cheriton, with whom I had worked closely at Arista. We saw a transformation in the networking industry and an opportunity to change the status quo. We saw the opportunity to provide the network engineers with tools that could, once and for all, empower them to deliver on their critical job functions and on the promise of a network that no longer stands in the way of the business.
Today is an exciting day. After two years of hard work, we are launching Apstra. We can finally tell the world about our mission and the Apstra Operating System (AOS), the product that our customers are calling game changing.
Reflecting On the Past Five Years
Whether your business is Software as a Service, banking, pharmaceuticals, or transportation, your data center network is the most critical foundation of your business success. Yet, having worked with network engineers first hand, it is clear to me that their requirements have not been met. Network Engineers are still managing their networks manually box by box; they are unable to use the hardware of their choosing; and they still lack the tools to prevent or debug outages effectively.
Because they don’t have the modern tools that enable them to keep pace with the demands of the business, network engineers are under extreme pressure and CIOs are dealing with network infrastructures that do not support their business goals effectively.
With SDN came a promise that all those problems would be fixed, yet these early approaches either argued that the physical network “didn’t matter” or that the protocols network engineers have used for the past 20+ years were the problem and needed to be replaced by some new magical protocol. Practice demonstrated that, despite all the hype, none of these approaches panned out and the network engineers were left holding the bag.
Network engineers are frustrated, their users are frustrated, and businesses are seriously looking to change.
Charting a Better Path Forward
In more recent years, we have witnessed an unprecedented shift. Innovative customers have led a transformation that finally has the potential to fundamentally change how they build and operate their networks. They are demanding physical and virtual networking devices from leading vendors to be programmed through published APIs. CLI wars are coming to an end and the landscape has shifted to a battle of APIs. Customers have also pushed vendors to recognize the visibility problem and provide access to real-time, high resolution telemetry from their devices. Additionally, the number of hardware choices available to them has grown steadily, which gives customers more leverage in ensuring that their demands are met. The ability to take advantage of these choices is now more important than ever.
Still, network engineers face several hurdles to take advantage of these new developments. Network engineers should not have to become programmers just to operationalize their networks! They should not be required to understand the subtle differences between vendor APIs and their semantics. They should not have to build automation systems using scripting tools that were designed for server automation. The network is a distributed system, and pushing different configurations to hundreds if not thousands of devices, each with a different role in the network, without any feedback can lead to serious errors.
This tectonic shift means that the timing is finally right to build the product network engineers have always wanted. This is the mission we have undertaken at Apstra.
The Apstra Solution
In laying the foundation of the Apstra solution, we recognized the following:
- The key to simplifying operations is to run the network as a system, as opposed to box by box. Networking systems are a distributed set of equipment running distributed protocols and routing applications. Network engineering requires complex control and visibility of this distributed set of equipment. Therefore, a distributed systems approach is required.
- An intent-driven approach that focuses the network engineer on the services they require from the network is key to managing the network as one system. Starting with a declaration of intent, network engineers need an integrated solution that automates all aspects of Day 0, 1, and 2+ operations with closed-loop continuous validation at the core.
- Network Engineers must maintain choice and control of the network equipment suppliers. They cannot afford to lock in their data center operational model based on the hardware vendor they happen to have deployed in their networks. Also, they need to stick to a horizontally layered architecture as much as possible – leveraging the protocols that served the industry well for the past 20 years is essential.
We are excited to present The Apstra Operating System (AOS). AOS takes a distributed systems approach, is intent driven, and integrates all aspects of Day 0, 1, and 2+ operations including closed loop, continuous validation. Being intent-driven, it takes a device-driver approach to supporting a variety of vendor equipment and works in conjunction with the per-box network operating systems.
AOS is truly vendor-agnostic as it enables network engineers to operate their network as one system independent of the different device vendors in the network. AOS delivers on three major goals:
- Operational simplicity and agility through powerful intent-driven automation.
- Operational control and visibility as network engineers are alerted to any anomalies that affect business delivery through real time telemetry.
- Massively reduced Total Cost of Ownership through powerful automation that enables hardware choice. Network Engineers are empowered to choose the hardware that best fits their business.
With AOS, organizations can now deploy or extend network services rapidly to support new business services without risk.
A Great Journey Takes Great Passion and Commitment
Two years in, we have assembled a phenomenal team of software and networking engineers at all layers of the stack — experts in physical devices, networking, distributed systems, abstractions, automation, user experience, and visualization — with a singular focus on our mission. We have been responsive and have iterated with customers quickly, incorporating customer feedback into the core of our solution. We have been working closely with customers, first sharing demos in 2014, then conducting trials in August 2015.
Customer response has been tremendous. Network engineers and CIOs get it. Our timing is spot on. David Cheriton’s commitment as a founder and investor ensures our independence and enables us to take the long-view, while being laser focused on customer needs and customer success.
We are at the beginning of a journey. As the father of 7-week-old and 5-year-old daughters, I often find myself making the analogy between software development and child development. You cannot rush child development, in the same way you cannot rush software; you need to give both the time to develop at their own pace to reach their maximum potential. We have had the luxury to do this. With AOS 1.0, we are starting a new chapter in the networking industry with the opportunity to challenge the status quo and think differently about managing the complex distributed system, which is the network infrastructure. Exciting times!
Empowering network engineers and taking control of the data center network are key prerequisites of business success. At Apstra, we are passionate and committed to help you achieve this goal. We invite you to join us on the journey.