Arista introduces EOS+

“As part of our work to connect billions of people around the world, we are building a network infrastructure that is more flexible, more scalable, and more efficient than almost anything else out there,” said Najam Ahmad, Vice President of Infrastructure at Facebook. “Arista EOS has proven to be a valuable component of our current designs, providing us with a series of useful features, including better control-plane and data-path programmability, the ability to write traffic steering and monitoring applications that integrate with Sysdb and the entire EOS stack running on our Arista devices, and an SDK framework is fairly easy to develop and test our code in. All this allows us to have more visibility in and greater control over our network — and that helps us continue to move fast as we scale.”

The Convergence of DevOps and NetOps
In the past, the networking, storage, compute, virtualization and applications worlds were built in silos. A more integrated approach results in organizational efficiencies and a much lower TCO for enterprises. The EOS+ platform allows compute, storage and application teams to integrate with the network, leveraging the Linux and programmatic foundations of EOS. Provisioning, monitoring and dynamic reconfiguration based on application workloads can now be integrated with the network in a programmatic fashion.

“The EOS SDK is a powerful platform for integration and scale, designed to strip cost and complexity out of network hardware. Working together with Arista EOS, Windows Server and System Center will help customers implement robust software defined networking from the switch to the operating system to improve datacenter operations and enable cloud flexibility,” said Albert Greenberg, Distinguished Engineer, Microsoft Azure, at Microsoft.

The EOS+ Platform
• EOS SDK – a development framework that allows native access to all levels of EOS for custom development and integration with forwarding and routing stacks that leverage advanced features such as MPLS.
• vEOS - a virtual machine instance of EOS that includes the same control plane and management plane as the physical switches.
• EOS Applications – Pre-built integration with technology partners and DevOps systems such as Puppet and Splunk for provisioning and monitoring.
• EOS Consulting Services – Professional services for development of customized solutions for network automation.

With vEOS, customers can develop and certify EOS in a virtual network environment with a diminished need for physical switches for test and certification. In addition, the EOS Consulting Services component provides customers with access to highly qualified Arista network developers to assist in implementing custom applications. This is a speedier and more cost-effective alternative to the legacy approach, which often spans many months and involves costly lab expenses.

EOS network applications enable easier programmability at all levels of the stack. Examples of EOS Applications include ZTPServer for rapid provisioning and Arista’s Network Telemetry Application for Splunk Enterprise, which provides customers the opportunity to tailor-fit their network to their needs, while also utilizing end-to-end provisioning and automation. These prebuilt applications provide complete solutions, all with 24x7, hands-on support from Arista’s Technical Assistance Center (TAC) engineers. The announcement furthers Arista’s commitment to creating a vast, open partner ecosystem that provides the best user experience for its customer base. There are 20+ Arista partners contributing to EOS Applications, with that number expected to grow quickly.

Customizations for Cloud Networking
Arista also introduced the EOS SDK for customers and technology partners who are looking for a “do-it-yourself” approach to applications that are custom-fit for their network environments. The EOS SDK allows customers to develop applications that integrate directly with the switch operating system. This approach can be used for customizing IP routing, protecting against DDoS attacks through selective workflow-based filtering and analyzing data for fine-grained visibility. This setup is ideal for large enterprises and cloud providers that require a high degree of workload mobility and workflow visibility along with the proven effectiveness and reliability of today’s EOS.