Subscribe to Blog Notification Emails

Latest Blog Post

Arista's Evolution to Data-Driven Networking

Jayshree Ullal
by Jayshree Ullal on Nov 2, 2021 2:59:10 AM

Arista’s EOS (Extensible Operating System) has been nurtured over the past decade, taking the best principles of extensible, open and scalable networks. While SDN evangelists insisted that the right way to build networks started with the decoupling of hardware and software in the network, manipulated by a centralized, shared controller, many companies failed to provide the core customer requisite in a clean software architecture and implementation coupled with key technical differentiation. This has been the essence of Arista EOS.

The Arista EOS Journey

All modern network devices and systems are implemented largely in a monolithic software stack. Arista’s state-of-the-art switch OS is built upon the programmable and high-speed gigabit-based packet forwarding engine, paired with sophisticated, programmable control and management planes to orchestrate the operation of the network as a whole. This differentiated architecture, based on a network-wide database (NetDB) has been transformational to customers and the industry at large.

Network device configuration and administration have no standards to describe how network state should be stored within a switch or a network of switches and routers, and no standards exist to describe how management of the switch should be done (e.g., network automation, analytics or availability for devices). Arista’s EOS, based on well-known cloud principles, is now installed all over the world from client to campus to data center to cloud networking. Arista has developed the core EOS architecture on the concept of generating and then centralizing all of the device state in the network with a publish/ subscribe model. This centralized state database enables rapid root cause detection on any network events, giving customers the ability to "look back in time" to understand the past state of the network. Arista EOS continues to be the most reliable network operating system while covering the broadest range of platforms across multiple network domains: campus, data center, cloud, WAN, etc. This transformation from PINs to PICs, along with CloudVision’s network-wide management of 100s to 1000s of EOS-based switches defined the 2010-2020 decade.

Arista EOS Stack Evolution

As we enter the next frontier of cloud-grade networking, Arista believes in continuous innovation around our core architectural constructs and we are pleased to introduce the next evolution of the core EOS Stack Architecture—the evolution of our centralized network database into a multi-modal and multi-tenant capable data lake—allowing for EOS to generate multiple data modalities and allow external data ingestion and enrichment to populate the EOS Network Data Lake (NetDL). With this introduction, Arista is enriching its state-driven EOS with a data store building block as shown in the figure below.

Timelinefor Blog

Arista AVA for AI-driven Assist

While the NetDL foundation brings great value for data networking, an AI-based expert system is needed to assist in more complex tasks such as security and deeper troubleshooting. Arista AVA (Autonomous Virtual Assist) imitates human expertise at scale through an AI-based expert system to automate complex tasks. AVA augments Arista EOS NetDL to provide real-time ground truth of the data and status of network devices, their state and behavior of packets. AVA’s expert system orchestrates an ensemble of AI/ML techniques on the data-driven network based on NetDL. AVA’s depth of tools addresses micro networking challenges, including quality of experience (QoE) management, proactive NetOps, and secure threat analysis via network detection and response (NDR).

With Arista’s powerful combination of EOS NetDL and AVA, our customers will be able to gather, store and process multiple modalities of network-generated and network-related data, providing the ideal foundation for AI/ML data generation. For instance, AVA uses a supervised ML algorithm to help network operators find subtle "grey failures” yet switches to using unsupervised NLP techniques to identify potentially malicious domains from web searches. This can also identify root causes of network problems, suggest remediations and finally, detect hard-to-find threats hidden in encrypted traffic.

The Arista Ecosystem

Arista’s data-driven networking is based on proactive platforms, predictive deployments and prescriptive insights to deliver on business outcomes based on human and AI heuristics. AI/ML enrichment and analytics using supervised and unsupervised learning modes can enable a broad ecosystem of vendor/partners to deliver market and customer-specific security, application and network performance analysis on top of Arista EOS-based NetDL stack and AVA expert insights. The transformation to data-driven networking feeding continuous awareness and assurance brings a single source of truth and business outcomes. It cannot be achieved by us alone and requires an intelligent ecosystem of partners and thoughtful architecture that shares, federates and ingests structured and unstructured data formats across client to cloud sources and real-time data driven networks. We are proud to collaborate with a rich and open ecosystem of industry leaders including: Equinix, Microsoft, Palo Alto, Red Hat, Slack, Splunk, VMware, Zoom and Zscaler. 

Welcome to the new decade of client to cloud data-driven networking!

References:

Opinions expressed here are the personal opinions of the original authors, not of Arista Networks. The content is provided for informational purposes only and is not meant to be an endorsement or representation by Arista Networks or any other party.

Jayshree Ullal
Written by Jayshree Ullal
As CEO and Chairperson of Arista, Jayshree Ullal is responsible for Arista's business and thought leadership in AI and cloud networking. She led the company to a historic and successful IPO in June 2014 from zero to a multibillion-dollar business. Formerly Jayshree was Senior Vice President at Cisco, responsible for a $10B business in datacenter, switching and services. With more than 40 years of networking experience, she is the recipient of numerous awards including E&Y's "Entrepreneur of the Year" in 2015, Barron's "World's Best CEOs" in 2018 and one of Fortune's "Top 20 Business persons" in 2019. Jayshree holds a B.S. in Engineering (Electrical) and an M.S. degree in engineering management. She is a recipient of the SFSU and SCU Distinguished Alumni Awards in 2013 and 2016.

Related posts

The New AI Era: Networking for AI and AI for Networking*

As we all recover from NVIDIA’s exhilarating GTC 2024 in San Jose last week, AI state-of-the-art news seems fast and furious....

Jayshree Ullal
By Jayshree Ullal - March 25, 2024
The Arrival of Open AI Networking

Recently I attended the 50th golden anniversary of Ethernet at the Computer History Museum. It was a reminder of how familiar...

Jayshree Ullal
By Jayshree Ullal - July 19, 2023
Network Identity Redefined for Zero Trust Enterprises

The perimeter of networks is changing and collapsing. In a zero trust network, no one and no thing is trusted from inside or...

Jayshree Ullal
By Jayshree Ullal - April 24, 2023