infrastructure orchestration and automation
As the recession continues to affect businesses across all industries, enterprise network resilience has never been more critical. The typical outage costs at least $100,000—a price tag that most companies can’t easily absorb in the current economic climate. However, decreasing business revenues have caused many companies, especially in the tech industry, to lay off large portions of their key IT staff. That means there are fewer administrators to monitor and manage network infrastructure and fewer engineers available to respond to issues and recover from outages.

Network automation is the key to ensuring 24/7 availability and optimal performance with less human interaction. A network automation framework provides all the tools and guidance needed to create a fully-automated network infrastructure that’s resilient to failure.

The four building blocks of a resilient network automation framework include:

  1. IT/OT production infrastructure
  2. Automation infrastructure
  3. Orchestration infrastructure
  4. AIOps

In previous blogs we discussed the role of IT/OT production infrastructure in network automation and how an IT/OT convergence strategy accelerates network automation. We also described the automation infrastructure components that enable end-to-end network automation. In this post, we’ll explain how infrastructure orchestration and automation build upon the previous two layers to enable streamlined, hyperautomated network resiliency. Our final blog in the series will conclude with a guide to using AIOps and other machine learning technologies to complete the network automation framework.

What is infrastructure orchestration and automation?

The infrastructure orchestration and automation layer contains the tools and paradigms used to efficiently manage and control that automation. The core components of infrastructure orchestration and automation include:

Version control

The automation infrastructure layer uses infrastructure as code (IaC) to decouple device configurations from the underlying hardware so they can be written as scripts or definition files that automatically provision network resources. In addition, this layer uses software-defined networking (SDN) to create a virtual control plane that overlays the production network infrastructure, allowing network management and optimization tasks to be written as automated scripts.

The goal of IaC and SDN is to reduce human error, speed up device provisioning, and build a more streamlined and resilient network infrastructure. However, IaC and SDN programming can be very complex, and not all sysadmins and network administrators are expert coders. In addition, an automated enterprise network has hundreds or even thousands of these definition files and scripts to store, manage, and deploy.

This is why a network automation framework should include version control in the orchestration and automation layer. Version control is a very familiar concept to programmers, especially in DevOps environments, but not all network and infrastructure teams have used it before. Version control involves storing all code in a centralized repository and then tracking and managing changes to that code.

Let’s say one administrator is responsible for configuring and maintaining the IaC definition file used to provision a particular model of Meraki AP. Here are some examples of how that workflow could break down when that one admin is out of the office for an extended period of time due to COVID-19 or gets laid off due to cutbacks in the organization:

  • Twenty new Meraki APs need to be deployed to a new site with identical configurations.
  • The existing definition needs to be updated and pushed out ASAP to patch a security vulnerability.
  • Someone discovers an error in the current version and they need to roll back to a previous configuration.

A version control system for IaC and SDN acts as the single source of truth for the entire automated infrastructure. All automation scripts and definition files are stored in one centralized location, so anyone with authorization can deploy identical devices with the push of a button. When an admin needs to change the code, those changes are tracked and can be rolled back at any time if a mistake is made. Version control systems even allow admins to leave notes explaining the reasoning or logic behind individual changes, so other team members can pick up where they left off, or in their absence, identify the root cause of issues.

Another key benefit of version control is that it facilitates the use of automated testing. QA and security analysts can run automated scans on code in the version control repository pre-production, so any misconfigurations or security vulnerabilities are identified and fixed before deployment. This reduces the risk of human error and improves the security and resiliency of the automated network infrastructure.

Version control is a core component of infrastructure orchestration and automation because it serves as the single source of truth for the entire automated network architecture.

Orchestrator

Automation is meant to make life easier, but it can be very complicated to manage on a large scale. Modern enterprise network architectures include thousands of moving parts in locations around the world and in the cloud. Automating each of these workflows means writing, testing, deploying, managing, and troubleshooting many different definition files and automation scripts. Doing all of that manually adds more work to overloaded and under-resourced network infrastructure teams, which increases the risk of something going wrong. Simply put, organizations need a way to automate their automation.

An orchestrator is a tool used to control all of the automated workflows on an enterprise network, just like a conductor orchestrates many different instruments and musicians into one cohesive symphony. An orchestrator uses management devices, like Gen 3 OOB serial consoles and SD-WAN gateway routers, to gain control over the physical and virtual network infrastructure. Administrators program the orchestrator to automatically deploy definition files or networking scripts (which it pulls from the version control system) in response to certain triggers. That means admins could potentially automate every step in every workflow, removing the need for human intervention and reducing the chance of errors.

Plus, an orchestrator can react to events much faster than even the best administrator. For example, if a spike in demand is overloading resources at one regional data center, the orchestrator can instantly deploy automated load-balancing workflows to reroute traffic before end-users notice any performance issues. This allows enterprises to maintain 24/7 network availability and performance even with reduced IT staff.

As part of a resilient network automation framework, the orchestrator should be vendor-agnostic (vendor-neutral). It needs to be compatible with all of the automation infrastructure components, as well as the production IT/OT solutions. It also needs to support all of the major third-party automation vendors, such as Ansible and Gluware, to give infrastructure teams the flexibility to use the tools they’re most comfortable with and that work best in their enterprise’s unique environment. Finally, the orchestrator needs to integrate with other tools within the orchestration and automation layer, including the version control system and the monitoring and analytics platform.

The orchestrator is what gives the “orchestration and automation” layer its name. It provides admins with the ability to automatically manage all the automated workflows that make up a resilient network infrastructure. An orchestrator reduces the risk of outages caused by human error and can automatically respond to and prevent potential issues.

Visibility & insights

It’s tempting to think of infrastructure orchestration and automation as a “set it and forget it” solution that can perfectly manage an enterprise network without any human oversight, but the technology isn’t quite there yet. Administrators need a way to monitor all the automated workflows, identify problems the orchestrator may have missed, and analyze the health and performance of the network infrastructure.

A visibility and insights platform collects logs from all the various components of the automated network infrastructure and aggregates the data in one centralized location. It provides visualizations of current device health and network performance, and may even include predictive analysis to power business insights. This gives administrators a big-picture overview of distributed, complex, and automated network architectures so they can ensure continuous availability and optimal performance.

As with the version control system and the orchestrator, the visibility and insights solution needs to be vendor-agnostic so it can dig into every single hardware and software solution in the automated network infrastructure. In a resilient network automation framework, the vendor-neutral version control, orchestrator, and visibility solutions are all combined in a single platform.

Infrastructure orchestration and automation with a single platform

A unified infrastructure orchestration and automation platform like ZPE Cloud simplifies the control and management of a fully-automated enterprise network. ZPE Cloud uses Nodegrid hardware—such as Gen 3 OOB serial consoles and integrated network edge routers—to deliver orchestration and automation to large, distributed, multi-vendor network infrastructures. The ZPE Cloud management app supports integrations with your choice of third-party version control and infrastructure automation solutions, or you can use Nodegrid hardware to directly host your automation software.

With ZPE Cloud, you also get comprehensive monitoring data on all connected infrastructure, plus, you can use Nodegrid environmental monitor sensors to gain insights on conditions in remote data centers and network closets.

ZPE’s Network Automation Blueprint

Infrastructure orchestration and automation works together with IT/OT production infrastructure, automation infrastructure, and AIOps to ensure network resiliency during uncertain times. The Network Automation Blueprint from ZPE Systems provides a reference architecture for achieving Gartner’s definition of hyperautomation as well as meeting the Open Networking User Group (ONUG) Orchestration and Automation recommendations.

In a future blog post, we’ll discuss the remaining building block of the Network Automation Blueprint in depth. In the meantime, you can read about IT/OT production infrastructure and automation infrastructure, or click here to get a sneak peek of the blueprint, which includes a 10-step checklist to get started with automation now.

Ready to learn more about infrastructure orchestration and automation?

To learn more about infrastructure orchestration and automation with ZPE Cloud and Nodegrid, contact ZPE Systems today.

Contact Us