The adoption of cloud services is a massive phenomenon that can deliver considerable benefits for telecommunications providers. These days, it’s hard to find a company that’s not running applications and workloads in the cloud because the potential business gains are too compelling to ignore.
Research firm Gartner projected that the worldwide public cloud services market would total $246.8 billion in 2017, up 18 percent from $209.2 billion in 2016. The highest growth will come from cloud system infrastructure services, which the firm projected to grow 37 percent in 2017. Cloud application services were also expected to grow significantly.
International Data Corporation (IDC) forecast that spending on off-premises cloud IT infrastructure will grow at a five-year compound annual growth rate (CAGR) of 12 percent, reaching $47.2 billion in 2021. Public cloud data centers will account for 80 percent of this amount, IDC says. Combined with on-premises private cloud, overall spending on cloud IT infrastructure will grow at an 11 percent CAGR and will surpass spending on non-cloud IT infrastructure by 2020, the firm says.
As with any type of technology deployment, however, a move to the cloud can come with downsides — notably higher-than-expected costs, delays in starting up services, and vendor lock-in.
For many telecom companies, open source technologies — specifically OpenStack — are key to addressing these challenges. An integrated OpenStack platform is deployable in hours instead of days, priced at a reasonable entry point for a telco’s budget, and is typically maintained as a truly open source platform without vendor lock-in.
Telecom companies are already familiar with OpenStack. A large majority (86 percent) of communications providers consider OpenStack to be essential or important to their success, according to a survey of 113 companies worldwide conducted by Heavy Reading and commissioned by the OpenStack Foundation.
Facing unprecedented traffic demands and rapidly evolving customer expectations, telecom providers around the world are speeding up their adoption of network functions virtualization (NFV) to increase network agility and mitigate costs, the report said. And OpenStack has emerged as the NFV infrastructure platform of choice. Already a number of telecom providers have chosen to implement NFV with OpenStack.
The flexibility and versatility of OpenStack as a cloud and NFV platform enables telecom companies to combine business and communications IT under a single technology, noted Jonathan Bryce, executive director of the OpenStack Foundation, in a press release. The organization is seeing “incredible progress and rapid adoption” of OpenStack by the industry, Bryce said.
In addition to NFV, telecom companies said they were already using or testing use cases with OpenStack for the Internet of Things (IoT) and for the future rollout of 5G-based services.
Telecom companies can benefit in a variety of ways by using OpenStack. For one thing, new services can be delivered to enterprise and consumer end-users much faster. This is extremely important in today’s fast-paced market, where being agile is critical to success. They can also benefit from the avoidance of vendor lock-in that can lead to higher costs and a loss of flexibility. Overall, telecom companies will also experience operational and software cost savings.
Service providers have a couple of approaches available to set up OpenStack, depending on their size and/or available resources. Tier 1 providers typically have the capacity to take the do-it-yourself approach, but most Tier 2 and 3 providers do not. In this case, the optimal approach is to work with an integrator that can provide a full turnkey OpenStack solution on a particular piece of server hardware.
Cloud provisioning is a complicated affair, and OpenStack is no different. Relying on a partner company to handle the integration portion can speed up service trials and eventual production deployments.
A typical approach is to use up to nine 1U servers to set up your OpenStack controllers and tools, and then attach it to additional rack-level compute resources to run services across a set of virtual machines (VMs).
A more economical model is to find a system integrator that can supply a turnkey (pre-configured and pre-validated) solution using 2U converged hardware that features high-density modular compute nodes managed by redundant Gigabit Ethernet switches. The result means open clouds can be deployed within hours rather than days and are operationally more cost-efficient. These platforms can even include additional compute resources – all in the same 2U ‘cloud-in-a-box’ platform.
A key advantage here is a separation of hardware and software, whereby a telco operator can run multiple concurrent third-party virtual network functions (VNFs) on the same platform. So for example, the platform could be easily configured to support one VNF from Vendor A on node1 and another VNF from Vendor B on either the same node1 (but different socket) or other nodes in same enclosure.
There are a handful of OpenStack distributors in the market, each offering their own package and tools at different price points, and each with their own set of pros and cons.
However, what’s important is that the OpenStack package they buy comes with the promise of remaining truly open and being continuously maintained and upgraded. Whichever OpenStack environment is deployed, the operator should be able to add any third-party compute and storage resources, as needed.
While the bulk of the growth for OpenStack is in the public cloud, the numbers are increasing among enterprises and telco service providers, which are expected to even surpass revenues from public cloud implementations by 2021, according to a 451 Research report. There are good options that can help alleviate the burden of integrating open source software with white box or commercial-off-the-shelf (COTS) hardware. This ensures that private cloud builders — telco or otherwise — remain more focused on a newfound agility to deliver services that boost revenues and incur lower infrastructure costs.
Comments