Sunset Learning Turns 30! Discover our story, exclusives, and celebrations →
→
From Dread to Excitement: Better Networks with Arista Automation
By Tony Bourke | 60 Min Video
Network automation can feel intimidating, especially when managing complex Arista networks. But with the right tools and guidance, it can actually make your job easier, faster, and more reliable.
Tony Bourke demonstrates how Arista solutions like EVPN/VXLAN, CloudVision, and AVD empower network administrators to streamline operations and scale with confidence. Learn how automation can transform your daily workflow.
In this video, you’ll learn how to:
- Generate and deploy configurations efficiently across Arista devices
- Test operational state and reduce errors with automation
- Leverage Ansible, Python, and Arista tools for real-world network tasks
Transform your Arista network operations from tedious to seamless, and gain the confidence to take automation to the next level.
Watch more videos like this on our YouTube Channel.
Automation Scale and Efficiency
- Arista AVD generates 40,000 lines of EOS config from only 4,000 lines of YAML, reducing manual configuration effort by 90% while eliminating copy-paste errors that cause outages.
- AVD builds, deploys, and tests a complete 48-node EVPN/VXLAN fabric with MLAG leaves, VLANs, and VRFs in approximately 30 seconds total (10 seconds for config generation, 20 seconds for testing).
- Changing the entire fabric’s underlay routing protocol (OSPF, ISIS, or BGP) requires modifying just a single line in the YAML data model, demonstrating extreme flexibility in network design.
Testing and Validation
AVD automatically generates 1,200 post-deployment tests for a 48-node fabric, validating EVPN peering, loopbacks, and vTap presence across all devices with zero failures in 20 seconds.CloudVision provides pre-deployment validations, rollbacks, and change control integration, offering safety features beyond open-source tools like Ansible or custom Python scripts.
Day 2 Operations
Adding VLANs and port profiles requires only editing YAML files, enabling rapid day 2 operations like port changes, QoS, and multicast configuration with minimal effort.
CloudVision enables rollback of the entire fabric to any previous configuration within the last 90 days, providing rapid recovery from deployment issues.
CI/CD Integration
Jenkins CI/CD pipelines automate the complete network build, deploy, and test process, integrating with Git version control for branching and pull requests to ensure team collaboration.
Architecture and Capabilities
- AVD supports building complete network architectures, including layer 3 leaf spine, EVPN/VXLAN, and layer 2 leaf spine topologies from YAML data models.
- AVD generates 30-40k lines of configuration in 20 seconds for a 48-node fabric, handling complex requirements like stretched VLANs across the entire topology.
Operational Philosophy
AVD’s YAML data model abstracts network configuration, allowing engineers to focus on design intent rather than manual CLI work, with Ansible serving as the deployment engine. Network automation addresses the dual challenges of low confidence and high cost of failure in traditional networking by eliminating human errors from typos, pasting mistakes, and manual configuration through extensive automated testing.