For the “next big thing in the data centre” software defined networking (SDN) has been keeping awfully quiet. So much so that 40 per cent of respondents to a recent Computing survey of UK data centre operatives claimed they’d never heard of it (figure 1).
This will be unwelcome (but perhaps not surprising) news for the networking vendors and start-ups who are positioning SDN as the logical next step in the data centre virtualisation process and who have been touting its benefits for some years now.
So what are these benefits, why aren’t vendors shouting about them more, and why are they – so far at least – being ignored by data centre managers?
With SDN the control plane of the network is abstracted from the physical switches and routers by virtualising it and implementing it as software. This leaves the forwarding plane dedicated solely to moving traffic around the network.
Abstracting the management of the infrastructure in this way makes for a much more scalable and flexible network, allowing changes to be made to the network topology dynamically, without the need to over-provision ports, as is common with traditional network architectures.
Traffic management and security policies only have to be defined once, making for simpler administration, and common management tasks can be automated. Also, by allowing visibility of network traffic at the application level the network is made application-aware, enabling finer control and automation of the business logic used to manage network structure and behaviour.
In short, with SDN the network moves from being a static monolith to a much more flexible backbone, able to adapt dynamically to changes in demand and providing end-to-end visibility.
That, at least, is the theory. But despite those benefits on offer only six per cent of our sample was seriously considering implementing SDN. Perhaps data centre managers (those who have heard of SDN) fear that it is still beset by conflicting visions, with standards, protocols and best practice yet to be properly defined. Indeed, 31 per cent of those familiar with SDN but not considering it in practice raised the technology’s immaturity as a sticking point (figure 2).
A similar number were not confident that they had sufficient skills in-house – not surprising given the tiny share of the networking infrastructure under SDN at the present time making it pretty much an unknown quantity. Similarly, issues of cost and uncertain return on investment will always be raised in relation to a new and untried technology.
Sixteen per cent thought it would add complexity, especially in a multi-vendor environment. This is related to the question of maturity and it is certainly something the industry must address. Recently, many networking vendors, including major players such as HP and Brocade, have lined up to support open technologies and standards, such as OpenFlow from the Open Networking Foundation – although some, notably Cisco, have not.
Others thought they would need to rip and replace their existing infrastructure to implement SDN, an argument that has an element of truth, although the same claim could be levelled at a conventional network upgrade. Moreover, the rejoinder from SDN vendors would be that, once on the SDN path, future changes should be much less disruptive. Similarly, any network upgrade can cause disruption and downtime, a risk that would, likewise, be greatly minimised post-SDN deployment.
It would appear, then, that SDN vendors face an uphill battle to persuade IT leaders that their technology is the answer. Few potential customers are familiar with its benefits and those that are fear that it is immature. However, they can comfort themselves that time may well be on their side.
Faced with the triple whammy of cloud computing, spiralling mobile demand and the incipient Internet of Things, IT leaders need to decide whether infrastructures designed for an era of more modest and predictable traffic growth are still up to the job.
In addition to the factors mentioned above, social media, video streaming, big data, BYOD and distributed computing are all piling on the pressure – pressure that will only increase over time.
These trends are becoming increasingly important for all types of businesses (and increasingly onerous for IT teams) and they are all completely dependent on the underlying network technology.
Up to now most data centres have been able to rely on regular increases in raw bandwidth plus advances in traffic-shaping and routing technology to keep pace with IT developments, but for how much longer will this be the case?
It is likely that some sort of step-change will be necessary to accommodate these changes. But for now most organisations assume that they don’t currently need SDN, with the current refresh cycle focused on upgrading traditional technologies in the traditional way.
When asked if they were planning to upgrade their network infrastructures to cope with increasing complexity and demand, a mere 17 per cent said they had no plans, with a third planning an upgrade within 12 months. Another 11 per cent expected upgrades to proceed within 18 months, with a further 21 per cent likely to follow suit thereafter.
The top two reasons for upgrading were the need to replace infrastructure at the end of its life, and the perennial problem of bandwidth. Better application performance was also cited. Support for converged communications was another popular reason for upgrading. Mergers and moving to larger (or in some cases smaller) data centres were among the other reasons why networks needed to be upgraded.
Whatever the reason, the general picture is one of widespread and ongoing infrastructure upgrades in progress across the whole corporate spectrum.
This picture begs the question why, when it would seem like the ideal time to, at the very least, evaluate SDN, so few companies seem to be lining up to take advantage of what it has to offer?
Just as virtualisation has revolutionised the efficiency and flexibility of server processing and the management of storage, it seems logical that eventually the same approach will become commonplace on the networks that connect servers, storage and clients. However, for now most organisations just don’t seem to see the need, preferring to stick with what they know. Apart from a few large data-intensive sectors such as telecoms, SDN has yet to make its mark in the wider corporate world.
“SDN is an interesting technology; there are engineering folks who might say that the technology had arrived before any use cases,” BT’s director of data centre strategy and platform, Adam Low, told the audience at last year’s Computing Data Centre Summit.
Perhaps that’s it. Clearly there is a need for more education and marketing efforts from the vendors involved, and for more use cases to emerge. Until then, SDN remains a case of “watch this space”.
Sometimes, the power of the mainframe is the most cost effective answer. Computing's Peter Gothard puts Computing's readers' questions on the future of the mainframe to IBM's Z13 expert Steven Dickens.
This Dummies white paper will help you better understand business process management (BPM)