Top Of The POPs
I always find it interesting when unanticipated themes crop up out of nowhere that are suddenly commonplace between multiple conversations with different vendors, especially vendors that are very much in complimentary, rather than competing, spheres of IT.
The latest such example is with respect to backbones and their POPs – AKA Points of Presence. So, sorry for all you pop-picker readers who thought I was somehow linking this blog to a classic BBC Top Of The Pops episode from the 70s or 80s. The PoPs and their interconnecting networks are of real concern when it comes to service delivery.
The basics are simple: we all know about fundamental Internet service delivery issues from where the closest ADSL/fibre wiring closet is to your endpoint, whether that service is from BT, Telefonica, DT etc etc. But in speaking with a number of different vendors I’m working with about this “everything as a service” era of IT, the common feedback they are getting from potential customers is that non-proximity to PoPs is causing serious performance problems, whatever that service may be. For example, it might be companies who have invested heavily in a SaaS platform or some form of network/security as a service solution, but we are talking serious OpEx investments with large-scale vendors here. In some cases – for example in the contact/call centre world, my VoIP and data optimisation client Aritari has been assisting with scenarios where voice and data performance missed the required mark when using multi-billion-dollar vendor SaaS solutions. Aritari found that the nearest POP delivering those services wasn’t even in the same continent!
The other issue – and one that is not so readily resolved – is when an entire service platform (SASE would be a great example) is underperforming and can’t be so easily “upgraded”. A random glance through any list of IT vendors nowadays quickly shows just how many are offering their solution “as a service” rather than simply the nuts and bolts for you to complete the deployment with – i.e., the IT world we used to know. The common problems are classic delivery issues such as latency – still the number one killer for real-time apps especially – and reliability.
The problem for a prospective customer in any PoP-related service delivery is understanding the exact nature of that delivery platform. Some vendors seem to be clearer than others on what they can deliver and where they can deliver it; others less so. What it all boils down to is that there is always far more complexity in selecting a vendor-provided service than you would initially anticipate – or want. But then, no one said IT was easy – otherwise we’d all be out of a job 😊