The Rise of the Public Cloud for Critical Service Provider Operations
Use of the cloud beyond expensive carrier-grade private clouds has been slow to take off in the telecom industry.
If we look at the enterprise IT space, widespread adoption disproves many of the myths that communications service providers (CSPs) cling to and that prevent them from widely implementing a hybrid or full public cloud strategy. However, in 2019 the marketplace has shifted, with AWS, Microsoft Azure and Google Cloud being familiar components appearing on progressive strategy presentations.
For some time now, leading telecom software vendors have been espousing the benefits of the public cloud as an option for OSS/BSS and SDN/NFV, while simultaneously trying to prove robust new cloud security paradigms are highly secure.
From an operational point of view, CSPs have been entrenched in cost-cutting exercises since the ARPUs for consumer contracts levelled out. The concept of the telco cloud was forged in the spirit of huge cost saving potential, while supercharging operational agility and speed. Meanwhile in real world, the use of private clouds, virtualizing servers and other telco equipment is simply not producing the cost savings required. So, let us look at the benefits of implementing a hybrid or public cloud strategy, and examine whether the technology is fit for purpose in live telco operations.
Where are the cost savings?
Procuring software in the OSS/BSS space, has in the past had a relatively definite CapEx number attached, but with a highly subjective OpEx value, depending on the variable costs of deploying, data loading operating and maintaining that software. So the total cost of ownership (TCO) is often an unknowable value when a CSP is in the procurement process. If you remove the costs of System Integrator (SI) installation activities, large license fees for databases to vendors like Oracle, additional equipment requirements, network resources, electrical power, extra staffing, you have taken away around 75% of the total OpEx. These are the realities of most cloud deployments, however in the public cloud CSPs are not paying to set up that cloud environment and are sharing resources in a pool with all other users, thus keeping some of those savings rather spending them on an alternative deployment method.
What about agility and ability to scale?
The ability to deliver OSS/BSS in a hybrid cloud model is increasingly being requested as part of operator RFP’s. This is primarily as it fits in with their overall transformation strategy for cutting out unused overhead, being able to create and launch new services prototypes in vastly shorter timescales and to rapidly scale up the digital services which appear to be the most successful. On premise deployments can do some of this, but a degree of future-proofing is going on here, and CSP’s are following the lead of the web-scale digital retail giants and enterprise IT players. Massive pools of compute and elasticity are the calling cards of services like AWS, along with the mantra of ‘only pay for what you use’, which fits this model perfectly.
Why have the security concerns gone away?
Security concerns have not fully gone away, but they are significantly less prevalent than they were five, or even two years ago. The majority of public cloud providers have been showcasing their enterprise grade security credentials through publishing their customer lists. Financial institutions, government departments, airlines, political parties, retail giants; all using public cloud, more than likely as part of a hybrid cloud strategy and storing data and systems that are subject to laws or regulatory rules in private clouds. And in 2019 we see many Tier 1 CSP’s joining this list: Vodafone, Verizon, Comcast and Telefonica joining the likes of Netflix, Uber and AirBNB as headline customers.
The nature of the hybrid cloud, will allow CSPs to pick and choose which components of their operational framework exist in the public or private clouds, to optimize their spending and maximum agility. It is important that the OSS/BSS systems can keep step with this shift, with common software foundations, leveraging microservices-based architecture and a full DevOps approach. This may be the point at which we see the big cracks appear in those monolithic, on-premise systems, designed in a different computing era.