Data-driven cloud networking uses an open approach to cloud networking around a single consistent software platform, the Arista EOS® network stack, and network data lake architecture (NetDL™), with the application of artificial intelligence and machine learning (AI/ML) to automation and security challenges.
Arista cloud networking extends the guiding principles of cloud-scale operators into a portfolio of cloud networking solutions serving data center, hybrid cloud, campus, wide area, and low-latency environments. These principles include the use of open APIs, programmability at every layer, cloud automation, self-service zero-touch provisioning, and a standards-based Universal Cloud Network (UCN) deployment architecture.
Cloud networking innovations started with a pioneering new software platform, the Arista Extensible Operating System (EOS®), which provides switching, routing, state-streaming and telemetry functions across all Arista platforms. Arista EOS established a new standard in networking for large-scale cloud operators, opened the door to the widespread adoption of merchant silicon hardware in networks, and provided dramatic decreases in deployment and operating costs while delivering unprecedented reliability for service providers, enterprises, broadcasters and others.
NetDL - EOS Network Data Lake
The Arista EOS network stack architecture provides a foundation for consolidation of streamed device state, telemetry, packet, flow, alert, sensor and third-party data into an aggregated Network Data Lake (Arista EOS NetDL™). Arista EOS NetDL consolidates diverse datasets required for effectively applying AI/ML methods in NetOps and SecOps environments, and it presents a single API surface for access to network and network-related data for enhancing Arista and third-party applications.
Attributes of EOS NetDL:
- Key to harvesting intelligent decisions and insights from all available data even when exposed and transmitted in massive amounts and in many forms and contexts.
- Engineered from the ground up on our existing EOS state-sharing and streaming approaches – NetDL establishes EOS as the foundation of an intelligent data-driven network and a data-centric network operating platform.
- Combines continuous large-scale data collection, pre-processing analysis, and enrichment using state-of-the-art machine learning and AI-driven correlation methods.
- Allows algorithms to find patterns to apply predictive analytics and present prescriptive solutions to complex problems.
- Enables an ecosystem of third-party applications and tools to feed into and consume enriched data.
Arista AVA - AI and Machine Learning
Arista Autonomous Virtual Assist (Arista AVA™) is Arista’s AI technology. Utilizing machine learning and other AI technologies, it augments all aspects of pervasive visibility, continuous threat detection, and enforcement. AVA is extensible across many other operational use cases. For example, AVA can address challenges in Network Detection and Response (NDR), Quality of Experience (QoE) management, and proactive NetOps. Combined with distributed network-wide state and telemetry data, distributed sensor networks, and third-party data sources in NetDL, it can drive automation and extensibility in network design to a new and unprecedented level and can dramatically reduce the burden of securing and supporting networks.
CloudVision® is Arista’s turnkey network operation, automation, and visibility platform. It provides rich functionality that is useful across the entire enterprise, using the same state-sharing architecture and APIs used by EOS NetDL. CloudVision provides leadership domain-specific features in a data center, campus wired and wireless, and cloud networks and helps enterprises to simplify network operations by breaking down traditional network management silos.
Arista AVA and NetDL also enable a cloud networking ecosystem of Arista strategic partners and ISVs to deliver market and customer-specific intelligent insights and solutions for automation via continuous integration pipelines, media and entertainment, cyber-security, application and network performance monitoring, and other categories.
Matt Leonard, Senior Director, Technology Partner Program at Equinix
Vasu Jakkal, Corporate Vice President, Security, Compliance and Identity at Microsoft
Lee Klarich, Chief Product Officer at Palo Alto Networks
Thomas Anderson, Vice President, Red Hat Ansible Automation Platform at Red Hat
Slack Technologies, LLC, a Salesforce company
Bill Hustad, Splunk’s vice president of alliances and channel ecosystems
Umesh Mahajan, SVP and General Manager at VMware
Velchamy Sankarlingam, President of Product and Engineering
Punit Minocha, Executive Vice President, Business and Corporate Development at Zscaler
- .EOS Architecture White Paper
- .CloudVision White Paper
- .Arista Advantage
- .AVA White Paper
- .Arista Q&A Document with Andre Kindness on Data-Driven Networking
- .Routing Architecture Transformations and Use cases
- .Solution Brief for IP Peering
- .Spotify’s SDN Internet Router
- .OpenConfig: the emerging industry standard API for network elements
- .VXLAN Pseudowires White Paper
- .Arista Networks & Nuage Integration
- .Solving the Virtualization Conundrum White Paper
- .VMware & Arista Network Virtualization Reference Design Guide for VMware vSphere® Environments
- .NSXTM for vSphere with Arista CloudVision® Environments
- .EOS CloudVision® & VMware NSX™ Solution Brief
- .Arista & VXLAN Technical Brief
- .VXLAN: Scaling Data Center Designs
- . Arista Universal Cloud Network Design Guide
- .Big Data Applications Design Guide
- .Data Center Interconnection (DCI) with VXLAN Design Guide