Telco on the Public Cloud: What to Watch Out For
April 20, 2022 - Joe Hogan
Plenty has been published on the benefits of hosting telco OSS/BSS functions in the public cloud so I would say there are a lot of things going for it overall.
As a senior technical leader in one of the most progressive CSP’s in North America put it, working with their Hyperscaler allows them to make all their cloud-based expertise available eliminating any need to invest in hiring relevant talent themselves. Looking beyond the basic hosting, it is important to consider the various tools which the likes of AWS (Amazon Web Services), Microsoft Azure or the Google Cloud Platform provide. Individual products and/or services like security, databases, analytics, AI, autoscaling – to name a few – are especially useful and comprehensive in their potential capabilities. Such is the case with AWS with over 200 cloud-based products on market. These tools are readily available to help build, deploy, and manage the network and IT services used to host on the public cloud. It is likely that we do not give much thought to the selection process of such services – conveniently provided and enabled speed and innovation, which is imperative in today’s market.
However, it may be worth the consideration for CSP’s to pause and reevaluate the path they will set themselves on at the innocuous juncture of choosing a set of services from public cloud providers. Given that one of the promoted key benefits of the cloud is openness, ease of integration with multiple different solutions, the cloud native elements and open APIs –the foundation of the public cloud. Often championed as the panacea to vendor lock-in, enabling easy movement between vendors due to the openness of the cloud native technology.
Although I am not entirely convinced on the validity of this claim. In building a telco OSS (Operation Support Systems) network orchestration using certain cloud services from a particular public cloud vendor, an AI suite or security solution, then later deciding on a public cloud provider instead may not be as seamless of a migration process as anticipated. An analogy may better encapsulate this – for any budding DIY enthusiast, the decision to buy the first cordless drill is a turnkey moment although that may not be fully realised at first since most models come with batteries. As simple of a decision as it may seem, a few months down the line – as your DIY ambitions and self-confidence grow – you start finding it necessary to include additional tools to the collection. Say, for instance, a cordless angle grinder or oscillating multitool for those highly specified tasks you may or may not undertake one day.
The options available to you online are endless, except now you are constrained by your previous buying decision. You do not want to overpopulate your kit, or duplicate product types, so you opt for choices that allow you to use things like battery packs on multiple tools – this may or may not be possible in certain cases. The key factor to consider is that manufacturers are not commercially inept – most of the leading power tool producers make their batteries so as not to be interchangeable with competitor products. As such, buyers are locked into a manufacturer from the initial purchase decision, irrespective of whether there is other tools with better features and/or greater value available to them. Thus, that initial purchase significantly influences future buying decisions.
This line of logic is also applicable to the choices made in using a lot of the very convenient products and/or services a public cloud provider offers as part of the hosting platform. The fundamental decision to utilise these may make it much harder to move the IT or Network services to a competing cloud provider in the future. So, in the rush to develop open architectures using cloud native software, avoiding vendor lock in so as not to repeat past shortcomings is worth considering as we move forward in 5G SA and telco in the cloud.