When one of the world’s largest providers of telecommunications services looked to develop a global cloud infrastructure they elected to partner with VMware and leverage a software-defined approach to its network, says Mark Reynolds, general business and partner lead at VMware Sub-Saharan Africa.
The company currently leverages an extensive global infrastructure and a Tier 1 IP backbone to deliver secure private networks, to businesses around the world. Its challenge however was to deliver a global cloud, that could leverage this extensive global IP fabric (network), as well as allow cloud resources to be extended beyond individual data centre boundaries.

Why? Because it wanted to allow enterprise-class customers to move applications and data on-demand anywhere in the global cloud network, without interruption, in the event of a disaster. A familiar scenario for many African telecommunications companies.

The company thought it would need to perform a complete overhaul of its existing infrastructure that made use of a combination of VLANs, a series of specialised hardware it had in place, as well as what was proving to be expensive, manual provisioning processes that delivered its network services.

Because of this infrastructure it was bound by a number of key constraints. It was unable to deliver key services such as virtual machine mobility between availability zones or on demand provisioning for virtual private clouds, as the inflexibility of its network was constraining it significantly. It was at this point that it came to VMware.

With its journey to the cloud now mapped out the company took a few key considerations into account. The first of which was to put a strategy in place to leverage network virtualization (software-defined networking) for its extensive network assets. It also wanted to increase operational efficiency, subsequently reducing its now expansive network hardware costs as well as eliminate vendor lock-in.

By taking the first step in virtualizing its network the company saw an almost immediate benefit. The organisation was hit by a natural disaster that affected much of the power grids that it fed off of, this then forced its hand in testing the merits of this migration. The result was that it was able to provide live migration of applications between its data centres ahead of mandated power conservation “brown-outs,” without service interruption or specialised hardware.

A key consideration for African telecoms giants that find themselves in power strapped environments where electricity grids are unreliable or simply inadequate.

Now that the key components are in place, the company is starting to see true benefits of working in the cloud. Its customers are able to develop, test and deploy complete application clusters (multiple application workloads that work together) and provide live migration of the entire cluster between data centres on demand.

What’s more, what was once a siloed environment made up of independent physical network environments, can now all be viewed as a single pool of network capacity. Not just single infrastructures. The virtual networks are also so secure that they are able to offer excellent flexibility, a key factor required for a dynamic, global cloud infrastructure. The virtualization of the network has also allowed it to take advantage of isolation between tenants in a multi-tenant cloud, what this means is that the company simply doesn’t need to add new VLANs.

The above highlights the true power of network virtualization and how it can be used as a starting point for cloud integration across multiple sites particularly for large-scale organisations that find themselves in different geographies. The ability to see, and make your network act as a single resource, through network virtualization or software-defined networking, is a key first step in moving the large enterprise into the cloud.