Skip to content 
Introduction
The advent of 5G technology has marked a new era in the telecommunications industry. The promise of high speed, low latency, massive device support and massive network capacity has driven new heights of innovation and development. With increasing demand for better network performance driven by data creation, the Open RAN (O-RAN) ecosystem has emerged as a key factor in the 5G landscape.
O-RAN stands for Open Radio Access Network and is an initiative to create a more open, flexible, and interoperable architecture for the radio access network (RAN) in wireless communication systems. As a new generation of disaggregated cellular network infrastructure, it offers a more flexible, cost-effective, and scalable alternative to traditional proprietary radio access networks (RANs). It is also poised to domesticate the supply chain relative to telecommunications infrastructure in the United States – a desire given it is deemed as critical infrastructure.
In this blog, we will dive deep into how Anuta Networks ATOM enabled a Fortune-500 company and new mobile network operator (MNO) based in the U.S. to deploy an Open Radio Access Network (O-RAN) as a service to accelerate its nationwide 5G roll-out.
Customer Key Challenges
- Deployment and activation of a Radio Access Network (RAN) of 50,000 cell sites within a year
- Tedious manual provisioning was not feasible due to its strict time constraints and requirement for FCC minimum coverage to avoid massive penalties
- A highly complex ecosystem with multiple vendors providing different services such as transport devices, bare metal servers, workloads, ticketing and inventory systems, IP Address Management (IPAM), password management, Slack, and more
- Lack of visibility and dependency as a result of legacy siloed systems
Empowering Multi-Vendor and Multi-Domain Deployment with ATOM's Vendor-Agnostic Platform
To address key challenges faced by this and many other MNOs, let’s examine the requisite requirements for a Zero Touch Provisioning Orchestrator (ZTPO) platform and how ATOM has successfully met these requirements.
Support for Multi-Vendors and Multi-Domain Environments
As a vendor-agnostic platform, ATOM helped deploy Cisco and Juniper routers along with Dell and Supermicro servers at the cell sites. A typical automation tool supports automation either on network devices or on bare metal servers but not both. However, ATOM supported automation on both network devices and bare metal servers, the feature being one of the core strengths of ATOM as a ‘platform’ where a domain or a vendor is not a constraint as long as that vendor provides an API that supports automation.
ATOM extensible workflows made integration with external systems easy in this regard. These systems include Infoblox IPAM, Ciena Blue Planet Inventory (BPI) management, Gitlab, VMWare TCA, AWS Secret Manager, Jira, ServiceNow, and more.
Cloud-Native with Massively Scalable Architecture
ATOM is built with a microservices-based architecture that delivers massive scalability and reliability. ATOM is also deployed as Docker containers installed in K8S within AWS while leveraging native AWS infrastructure components. Most importantly, ATOM’s cloud-native architecture supports horizontal scale. This provides consistent, ultra-low latency even when a massive number of devices are onboarded. ATOM also provides the flexibility to handle parallel cell-site provisioning at scale.
Geo-Redundancy and Active-Active High Availability
Given the fact that ATOM’s cloud-native architecture supports horizontal scale, it delivers constant latency even when devices from distributed locations are onboarded. ATOM also supports a multi-site distributed server agent model that offers Geo-Redundancy and Active-Active High Availability. This architecture is highly efficient, secure, and resilient while providing a single pane of glass for all availability zones- making management much easier. Additionally, ATOM supports in-line upgrades without downtime or service performance degradation, reducing the likelihood of subscriber churn.
Zero-Touch Provisioning
ATOM provisions devices smoothly without affecting downtime or performance through fully automated Zero-Touch Provisioning (ZTP). In the event of errors, ATOM offers an option for auto-retry or manual intervention. ATOM’s extensible dashboard and deployment reports also provide comprehensive views of the provisioning process, allowing for quick identification of any issues.
Support Role-Based Access Control
Multiple teams, such as market managers, operators, server teams, network architects, and more have specific responsibilities for performing functions that access ZTP. Hence, a tool should support access and web interface based on multiple user roles. ATOM’s support for role-based access controls allows different roles to be defined for each user group and assigned according to usage. ATOM also provides support for single sign-on (SSO) and integration with active directory.
Extensible Platform
ATOM provides an extensible platform using ATOM SDKs that help DevOps teams to develop new workflows or extend existing workflows quickly and easily.
Compliance
Another MNO requirement is that a given platform must be able to generate compliance reports. As ATOM supports configuration, software, and hardware compliance reports, it facilitates the generation and download of required reports.
ATOM ZTPO and Cell Site Provisioning
ATOM – ZTP Orchestrator
Workflow: ZTPO Cell Site Provisioning – All Stages
ATOM comes pre-loaded with provisioning templates containing all of the necessary information for setting up cell sites. These templates encompass the details required for configuring external systems such as Gitlab, Infoblox, AWS, Jira, ServiceNow, Blue Planet Inventory, and more. They also include information for provisioning servers, such as BIOS and firmware versions, ESXi images, file server details, NTP servers, DNS domain, and passwords. To accommodate various geographical regions, vendors, and other factors, ATOM maintains multiple versions of these templates to ensure global compatibility.
It’s important to note that cell site information is also stored in external systems such as Infoblox IPAM, where IP addresses and VLAN allocations are kept, and Blue Planet, where inventory details are managed.
Workflow: ZTP Cell Site Provisioning – Router, BPI, Infoblox, PXE, Server Provisioning Stages
In this illustration, a field ops technician starts cell site provisioning by plugging in gear, which triggers the router’s ZTP process connecting to the EMS system. IP allocation is obtained from Infoblox, and the generated configurations are applied to the router.
ATOM ZTP is notified at the start and completion of the router provisioning. In addition to the Site-ID, a successful notification includes site GPS coordinates, radio unit IPs, the server’s BMC IP, and the MAC Address required for server provisioning.
ATOM ZTP uses Site-ID from the notification to cross-check if this site exists in the Blue Planet Inventory (BPI) system and if the GPS coordinates of the site match. If the site passes the checks, ATOM fetches additional cell site information required at the later stages of provisioning.
In the next stage, ATOM ZTP creates DNS entries for the radio units, reserves IPs, and creates DNS entries required for the server provisioning in Infoblox.
Workflow: ZTP Server Provisioning
Bare metal server provisioning starts with the triggering of the PXE boot process. ATOM makes use of Redfish APIs to provision Supermicro and Dell servers. The server calls ATOM ZTP after the PXE process. The callback includes information like IP addresses, interface details, and more., all required at the later stages of provisioning. File servers distributed geographically host the BIOS, Firmware, and ESXi images required for server provisioning. The server next receives the file server details from the DHCP server – Infoblox.
After PXE Boot, the server provisioning goes through the following steps:
- Update BMC settings – Passwords, NTP, DNS configs, etc.
- BIOS Upgrade
- Firmware Upgrade
- Deploy ESXi hypervisor
- Update ESXi Settings – Passwords
- NIC Driver checks and Upgrade
- Server Inventory (CPU, RAM, Storage, etc.)
Workflow: ZTPO Cell Site Provisioning – vCenter, Node-Pool, DU instantiation stages
Once the bare metal server is provisioned and ready, and in the subsequent stages vCenter is deployed, node pools are created, and ESXi security configurations are hardened. This is followed by DU instantiation. All of these stages involve invoking various GitLab projects that, in turn, interact with VMWare TCA to accomplish the activity. ATOM ZTP generates the input files at every provisioning stage using the information fetched from the BPI inventory database and uploads these files into AWS S3 buckets. These files are fetched by the GitLab projects during provisioning. Some of the password details are finally fetched from the AWS Secrets manager.
Interaction between External Systems and ATOM ZTPO
ATOM provides REST endpoints, and webhooks, so that external systems can send notifications – and vendor-provided APIs are used to facilitate the process. For example, external systems like router EMS, bare metal server, and Gitlab projects send status notifications to ATOM webhooks. ATOM uses REST APIs provided by Infoblox, BPI, and Gitlab to interact with them.
During certain stages of server provisioning when required APIs are not present to achieve the objective, ATOM uses CLI to provision and fetch data.
Customizing Deployment Reports with Flexible Filtering and Download Options
Efficient reporting is critical for large-scale deployments. ATOM’s extensible dashboard and deployment reports provide both summary and detailed views to help track the progress of the different stages of provisioning at cell sites and isolate any issues quickly.
Beyond Capability: Exploring the True Value of an Automation Platform
Typically, automation starts once a network design is finalized and tested. It is extremely difficult to automate a solution that is deployed for the first time and for which the design is still evolving. Besides a reliable automation platform, having a proficient team that can swiftly cater to needs is imperative, especially when there are strict deadlines. In this blog, we demonstrated how the Anuta Networks team is leveraging ZTP to demonstrate remarkable agility in fulfilling customer requirements while adhering to designated timelines. Some deliverables were completed in less than two weeks, several were accomplished within a few days, while a handful were finalized within a couple of hours. ATOM automated test suites also ensured that deliverables were of the highest quality and free from regression problems accompanied by 24/7 customer support, ensuring top-quality results.
In Part II of the blog, we will explore ATOM’s deployment reports and dashboards, which allowed this MNO to quickly identify and resolve issues with increased visibility into the provisioning process. We will also examine how ATOM handled errors during the deployment process and the role-based access control required to ensure that only authorized users have access to specific platform sections.
To learn more, read Case Study.
Additional Contributors: Manisha Dhan
