Each New Year brings with it new puzzles to solve, and questions to answer. These are six topics that I’ll be keeping an eye on in 2016.
Geoff Huston’s 2015 address/routing reports and what they mean
Geoff wrote two reports earlier this year, which highlight the behaviour of the system of address management as a whole in two dimensions: who is getting the resources (and how) and how are the resources being used.
This is a continuing story of the end of supply, and adoption of processes, which formally we call transfers and informally we think may be leasing ‘under the hood’. In routing, the continuing signals of where addresses are being re-used feeds the same story.
Will these trends continue? Will the availability of addresses in the AFRINIC region feed into transfers to other regions? How sustainable is farming of the pre-RIR system blocks? Time will tell but it’s worth asking these questions to remind us of these issues.
The growth of containerization
Docker and Kubernetes represent the emergence of cloud as it always should have been: the capital investment in machine-room racks, and CPU cores and disk. And now, it is available to us to run high-level abstract services.
The dependencies are defined. The risks of opportunistically exploitable holes from other services are low given a Docker instance doesn’t have to run any standard daemons, unlike a full blown virtual machine and taken to the logical conclusion, the underlying kernel of the OS can be stripped of all but the essential system calls to support the function in the container.
But, there are costs in addresses, and inter-process communications. How is this going to play out?
The growth of DNSSEC
Trust in the DNS is now becoming the basis of higher trust in end services.
If we continue to grow DNSSEC, can we start to get rid of a class of attacks on the system as a whole – implicit in spam and phishing attacks?
The growth of IPv6 – will it break the magic tipping point in 2016?
For some providers, IPv6 is already over the tipping point. But for many more it’s still far off in the horizon.
What does it mean for us all, if we bifurcate into a dualstack-v6/v4 and IPv4-only Internet? Are we creating a second-class network for some?
Carriers will not recapture the vertical market with OTT, consumers have spoken
The assumption that ‘cash cows’ such as SMS and voice can continue to sustain the Telco model are looking pretty weak. People now routinely swap out a SIM for another because they can use over-the-top (OTT) services from Facebook, WhatsApp, Skype, or Google to support everything they used to depend on from their telco.
Regulation of OTT services isn’t going to fly. So the money is going to pass away from this segment.
How will the upgrade to 5G be funded, if the cash cow applications, which defined growth in mobile/cellular technology, aren’t there?
Equally, the re-capture of the unmanaged/unlicenced spectrum poses questions for intermediate service delivery. If telcos succeed in building out wide scale WiFi and services like VoLTE, what does this do for people using WiFi spectrum for point to point?
Do we have ubiquitous crypto on the wire yet?
DNSSEC secures the name. DNSSEC can secure the trust in the source of the name’s Internet services. But what secures the communications? How much will models of ‘HTTPS/TLS all the time’ work?
Hotel networks, airport networks and other constrained public networks have been relying on proxy/mediating services to limit use. Business intelligence has depended on packet inspection to understand what people are doing. Where are these things going when significantly more communication is encrypted end-to-end?
The views expressed by the authors of this blog are their own and do not necessarily reflect the views of APNIC. Please note a Code of Conduct applies to this blog.