If you are a small company with a small network and around 50 devices, the degree of network automation that you need could be quite minimal. All you need is to automate day-to-day manual mundane tasks with Chef, Ansible or Puppet. The automation is rudimentary and not fully fleshed-out. It works in silos with bits and pieces of automation sprinkled across your network.
If your network is bigger with a wide variety of multi-vendor devices, basic automation may not be enough. Integrating Python, Ansible with NetBox and other apt tools, you could manage to create a single source-of-truth. You can effectively manage device configurations, enforce some compliance policies, and simplify device on-boarding processes with the automation framework. Wielding this, you may even have some partially automated software image-upgrades on your devices. Your framework is also well armed to collect and display interesting information from your devices.
Your automation is at a much more advanced stage. It’s not a point-and-shoot framework. Instead, it’s strategic and you have a firm grasp of technologies. The framework is intelligent, transaction-based and stateful. Your framework- which already packs the capability to monitor and collect tens and hundreds of devices – is now strengthened enough to display interesting information, as well as to alert and notify you in case of any issues. The entire automation platform is customizable through APIs and extensible through SDKs. You can now integrate the platform to any of your ticketing/billing/ITSM solutions such as ServiceNow, Jira, or BMC Ready. You can even integrate these new contours with existing pointed automation scripts.
The network automation framework is future-proof and armed with cutting-edge technologies. You integrate workflow, approval flows, and multi-tenancy into your automation suite. You implement technologies such as intent-based networking and closed-loop automation to eliminate any human errors further while also enforcing compliance and configuration management. You can define intents declaratively – and the platform will convert it into actions.
With support for a large variety of devices across numerous vendors, you are liberated from all worries about low-level configurations – the automation framework takes care of all that legwork and hassle. You confine yourself to the upper-crust work – like defining and creating policies and methods of procedures. The automation gains derived here bolster you with extensive in-depth analytics to monitor your entire network.
Supports 150+ platforms across 45+ vendors. ATOM platform is horizontally-scalable and supports thousands of devices with a single instance.
Provides an intuitive graphical-interface to design, develop and execute complex or straightforward network operations and procedures.
Collects streaming telemetry and provides meaningful insights on the network. Collected data can be used to define metrics and corrective actions can be undertaken to preserve network state.
Enforces compliance around-the-clock by automatically detecting violations and taking corrective actions. The platform regularly archives and preserves configuration on devices
It depends on what level of automation you are currently in. To achieve complete automation, a beginner may take 2-3 years; while an Expert may take around 6-9 months.
As few as possible. It becomes increasingly difficult to use, manage and maintain a large number of tools. You should aim to reduce the number of tools by installing feature-rich solutions covering most of your requirements.
Smaller networks might be amenable to an internally-developed automation solution, but maintenance and future-scalability areas are still at risk. On the other hand, larger, and more extensive multi-vendor and multi-domain networks will benefit greatly from a packaged, microservices-based automation solution.
The following checklist might provide insight into determining the fit of a packaged solution:
1. Is there a need to automate more than 500 devices on a given network?
2. Are there more than 2 vendors in a given network?
3. Do network operations need to be tied to business operations? i.e. integration of network workflows with ServiceNow/Jira, etc.
4. Is a single pane-of-glass desired to manage a given network, or can it be managed with multiple, distributed interfaces?
5. Is efficient provisioning required to effectively scale an automation framework as a network grows?
6. Are network enhancement and future-proofing parts of a long-term objective?
Any automation framework becomes sustainable, long-term and advantageous only when the right vendor supports it. You do not have to look for the brochures but for the outcomes.
Look for the following while selecting the automation vendor 1. Ease of doing business
2. Expertise in automation
3. Flexibility in pricing
4. Testimonials and reviews
5. Depth of post-sales support