Brian Daugherty, Senior Staff Consulting Architect; Timi Laszlo, Technical Product Manager & Adam Rabin, Product Marketing Manager
Next-generation network services like 5G, vRAN, private network and multi-access edge computing (MEC), will be crucial in the coming years to enabling new technologies. Solutions and opportunities such as supply chain optimization, autonomous driving, smart cities and IoT rely on a rapidly expanding telco cloud capable of handling many dynamic workloads all at once. These new services also require hundreds to thousands of near/far-edge sites and, currently, operators have limited options to efficiently provision and scale those sites en masse.
With vRAN set to become 20% of the addressable RAN market by 2025, it is no surprise that automating infrastructure deployment and management will play a key role in any CSP’s operational best practices.
Understanding the evolving needs of CSPs and their corresponding investment in new 5G-based solutions, VMware developed Infrastructure Automation (IA). VMware’s IA is the fourth layer of automation woven into VMware’s Telco Cloud Automation product—to facilitate rapid network expansion and support CSPs in their 5G and vRAN journey. VMware’s IA allows for rapid multi-site provisioning—improving an operators’ ability to deploy 5G faster and at lower cost.
How it Works
As demonstrated in this short demo above, an operator can simplify network expansion for core, edge and RAN use cases while also managing the complexity and cost of deploying several new sites. The demo presents three simple steps:
- Configuration of global settings applicable to every site;
- Domain specific configurations for central and regional sites; and
- Host configuration and provisioning for each new site
With this easy three-step process, an operator can configure the deployment of virtualized network infrastructure (SDDC) across all domains—from central and regional sites to compute and vRAN sites deployed at the edge. Once provisioned, it can then easily deploy workload-optimized Kubernetes Clusters and instantiate network functions or services at each site. The diagram below illustrates an example workflow in IA:
Benefits of Infrastructure Automation
IA is closely coupled with TCA CaaS automation, xNF management and network service orchestration for end-to-end automation. It is managed as a unified “product” contrary to other approaches that require integrating several tools to deliver similar capabilities. As such, IA offers multi-site management UI and a simple, uniform API ready for the operator’s CI-CD.
Among other benefits, IA enables
operators to deploy virtualized infrastructure and provision new sites easily
through automation and templates—thus diminishing the need for complex manual
site operations. As a result, IA not only reduces the potential financial cost related
to deployment efforts, but also speeds the provisioning of new sites, eliminates
the potential for human-error and misconfigurations leading to operational
inefficiencies or security vulnerabilities and has built-in auto-recovery from
possible network failures. Once initial sites are provisioned, IA then promotes
repeatability and scalability in constructing a network: operators can define,
in a declarative manner, how sites operate and then reuse those same site settings
for other ones.
IA is an important and necessary capability for
any operator seeking efficient site provisioning, powerful virtualized
infrastructure deployment and general flexibility in network management and
operations. In today’s 5G-centric world, speed and efficiency are paramount if
CSPs wish to unlock 5G and edge-based use cases for customers. Without using
IA, it might take an operator 10, 20 or even more hours to deploy a single site.
With IA, an operator can deploy a site in less than 2 hours—representing an 80%
or greater reduction in time to deploy new sites. IA is, therefore, crucial for
any operator seeking new 5G-related revenue opportunities and cost-cutting
 See: Analysys Mason report, “vRAN: the high-level implications for OSS,” by Anil Rao, April 14, 2020.