Intent-Based Data Center Automation

Software solution

Many of the network processes such as managing, creating, deploying, operating, and troubleshooting a data center network can be very difficult, expensive, and resource intensive. Today, The simplification of the IT infrastructure according to the market research and analysis results by Gartner the number one strategic priority for companies.

Apstra follows the intent-based networking paradigm, an approach in which the networks as an overall system and not as their individual components must be easier to manage. Furthermore, the success of a network should be defined by the results that the entire system achieves. The network should always fulfill this "intention" of the network architect. As an intention-based distributed operating system, AOS enables a logical model for the network to be created using computer algorithms based on a series of information or answers to specific questions. The necessary input from the architect / operator should be as minimal as possible. The required service can be directly communicated to the network without having to carry out every single task. It automatically generates and implements the complete configuration of all devices based on a service description and offers ongoing checks for deviations.

Apstra Operating System (AOS)

The Apstra Operating System (AOS) is a compact, cross-vendor overlay solution that automates the lifecycle of end-to-end network data center services. With the aim of cost reduction, the system allows network engineers to plan and manage the setup and operation of a spin-leaf network quickly and reliably. Suitable for networks of any size, it can be scaled to the largest data centers.

How does AOS works?

AOS already works with your existing hardware (e.g. with switches from Cisco, Juniper, Arista, Mellanox) and the operating system software (such as NX-OS, Junos, EOS and Cumulus). Brite box solutions (such as Dell EMC) or white box switches on which open source network operating system (NOS) runs are also supported. AOS completely separates the design, implementation and operation of your network from the physical (underlay) and virtual (overlay) network infrastructure, with each hardware and NOS you select being supported.

Apstra AOS - manufacturer independent Overlay

With simple buttons, the network design, the construction and provision of two servers up to thousands of racks can be automated in real time. It also enables troubleshooting, validation and policy translations, and self-documentation. With intent-based analytics, extended customization options for system tests can be carried out. In the event of network problems in the data center, you can also use Apstra's central support. This should make it unnecessary to receive new manufacturer certifications in order to operate alternative operating systems and network devices.

Function in Detail

Apstra AOS - Lifecycle

As a pure software implementation, the Apstra AOS operating system is installed on a server or on any VM. The device agents, which are installed on the device itself or outside, ensure that your existing network devices connect to the software. They can then be combined in the network design phase as logical devices and racks to form a template (blueprint). Further details such as Rack types, routing guidelines, EVPN / VxLAN, IPv4 / v6 and other resources are assigned to each template and assigned to a compatible physical device profile. Multiple templates can also be created and instantiated so that they can be created as a single design for the entire network. The configuration for the network generated from the blueprint can be easily changed using a repository function and can be played back in a few minutes in the event of errors. Thanks to the central management of the entire network via AOS, the network can be enlarged or reduced more easily, meaningful device telemetry extracted and an exact status of the network elements provided.

Advantages of AOS

The most important and innovative aspects of Apstra AOS software at a glance:

  • IPv6 support also for virtual and containerized computer environments.
  • Automated system monitoring and independent troubleshooting.
  • Free hardware / manufacturer and operating system choice that meets your business needs.
  • Apstra support for planning, initiation and deployment.
  • Management of thousands of devices as a single logical element, with a complex EVPN overlay, with just a few clicks of the mouse.
  • Simplified scaling for massive cloud and compute farms with over 100,000 servers.
  • Add racks or topology to existing pods for quick network expansion.
  • Expand, downsize and move network and servers without affecting business applications.
  • Network Lifecycle Day 0, 1, 2+ automation (move / add / change).
  • Seamless integration with VMware NSX, vSphere and Nutanix.
  • Group-based guidelines for easy management of network access for all manufacturers.
  • Extended cause detection & improved MTTR (Mean-Time-to-Repair).
  • Rapid update of the network operating system.
  • Intent-Based Analytics (IBA).
  • Datacenter Interconnect: EVPN overlay according to industry standard.
  • Automate the entire lifecycle of network design, deployment and operation.
  • Very easy to use and quickly adaptable.
  • Simple user interface, browser-based administration tool and open APIs.
  • Does not require complex programming.
  • Quick to use and adaptable to business requirements.
  • Reduces the time for operations up to 90%.
  • Eliminates up to 85% of all failures.
  • Reduces operating costs (up to 83% OpEx and 60% CapEx).

Are you interested? Contact us for more information.