It’s a challenging time for telcos, as the expectations of their subscribers rocket, and they attempt to mitigate a huge transition under the auspices of new networking and communications standards, writes Nathan Rader the director of NFV strategy at Canonical, the company behind Ubuntu.
The advent of 5G, and rising demands on both capacity and data are transforming the space beyond recognition. Early to shift to scale-out systems, telecommunications providers are deploying infrastructure on a large scale. Combine this with their focus on differentiating themselves in the market via cutting-edge services, and the outcome is a variety of very complicated platforms which can prove to be extremely expensive to maintain. For the majority of operators, even simple issues, such as the construction of functional and resilient technology for the cloud, can cause serious problems.
This leads to big questions around how telcos can make the economics of the cloud work to their advantage. In this article we’ll discuss three things telcos must consider to maximise their investment in the cloud.
A vendor agnostic path
Telcos have many highly complicated demands on the cloud, which means that infrastructure is usually expensive and very difficult to manage. This often means that telcos will choose pre-packaged stacks as they are the most straightforward and affordable solutions. While the lure of discounts that can be offered in purchasing these platforms is strong, it can lead to telcos being locked-in. When operators choose to put themselves in this position, they become vulnerable to unforeseen price hikes and a dearth of functionality later on, which could impinge on both growth and innovation. When locked-in, changing any part of the stack without reprogramming or implementing new solutions for the rest is much more problematic.
This makes choosing the best mix for the task at hand – OpenStack cloud and NFV solutions – is key for telco success. CIOs must look for open source operating systems which need a lower amount of proprietary system development, but also unlock full flexibility with large ecosystems, personalisation, high support levels and security. Intrinsically, this means choosing a range of varied solutions from a pool of vendors and uniting them with the correct open source platform to streamline deployment and management. It is absolutely crucial to select solutions which are capable of continually evolving and automatic improvement in tandem with the development cycle. Looking ahead, working in this way supports lower costs and uncovers the most suitable platform for innovations and scaling.
The automation of operations
Industrial scale automation is the only feasible way to manage the network environment cost effectively. Amid fierce competition, telcos are always seeking to differentiate themselves technically. However, the constant adaptation of software for individual purposes produces ’snowflakes’. These complicated and tailored services depend on the systems architects that created them staying in-role. This makes it nigh on impossible to automate each upgrade and maintenance cycle. Each snowflake requires special configuration away from the automated deployment scripts, which draws both time and resource from developers. Fully-automated DevOps pipelines work much more effectively when each piece of software is the same, enabling telcos to roll out thousands of identical operations with half the time and trouble, rather than running a handful of deployments which need manual fine-tuning.
Deployment modelling
Telcos are under pressure to increase their network capabilities and capacity across developing markets. To keep pace with this kind of growth, telcos must be capable of duplicating the environments they have in current markets in a simple way. When well-supported models have been implemented, it is easy to overlay more personalised functions, or services, over the top of other models to meet the exacting needs of each region. An automated infrastructure can be packaged up in containers and redeployed to servers spanning the globe, enabling telcos to slash latency in multiple regions.
Getting the house in order
While the lion share of operators are planting the flag on advancements such as 5G and edge computing to boost differentiation, they will not to able to enjoy the advantages either offer if they don’t organise their software defined network (SDN) and network functions virtualisation (NFV) infrastructure. For example, contemporary NFV solutions that enable traditional and modern real-time communications can still fall prey to jitters and delays. They need to be deployed in a way that doesn’t eat up the systems resources prior to scaling to a size that delivers a good return on investment.
The telco industry is under deep pressure to accelerate scalability and on-demand computing, reduce infrastructure and operational management costs and shorten the time it takes to get new applications to market. Possessing the correct tools and operating infrastructure is crucial to realising the value of investment in the cloud.
