• 12/12/2013
    4:30 PM
  • Rating: 
    0 votes
    Vote up!
    Vote down!

7 Ways Networking Will Evolve In 2014

Network infrastructure is experiencing some dramatic changes in response to what's happening in the IT landscape. Here's what to expect in 2014.

There is a fundamental shift happening in information technology as it moves toward software-based and services-led architectures, and this trend will resonate more than ever in 2014. This translates into some significant changes in how we evaluate and deploy networking functions. Join me as I outline the IT networking concepts that will take shape in 2014.

1. The hypervisor-centric SDN model will reign supreme.
The hype and interest around software-defined networking is at an all-time high, and while the industry continues to sift through the endless marketing messages and potential deployment options, I believe we will see hypervisor-based SDN models win out over switch-based SDN for seamless control of the network. Many IT departments don't have the money or resources to forklift their existing switching and routing infrastructures. And while vendors will continue to get wrapped around the axle with open standards, we will see more vendor-driven SDN architectures, which will continue to win out over the pure open standards.

2. Virtualized datacenters will enable new Layers 4-7 service-oriented deployment models.
The increased adoption of virtualization in the datacenter will result in more point-and-click service enablement directly from the virtual domain. This new model of service delivery will rely less on complex network configurations and expertise. By separating network services from physical equipment and underlying network infrastructure, these services can also be abstracted and orchestrated as a common pool of resources for end-users.

As such, the tools for provisioning, configuration, and management of those services can be radically simplified and automated. Network services, once exclusively the domain of networking specialists, can be virtualized and off-loaded to virtual administrators and application owners responsible for an application's experience. The combination of these benefits makes network virtualization -- and virtual network services in particular -- essential components to realizing the software-defined datacenter.

3. Networking as a standalone job function will come under attack.
Virtualization and SDN have started to make legacy networking expertise irrelevant. A next-generation IT organization will start to develop. These organizations will look for diversified skillsets that combine networking, storage, and virtualization expertise. We are beginning to see the early stages of IT organizational shifts happening today. Functions that were once the sole domain of the network administrator are now becoming part of the virtual administrator. And network administrators looking to deploy services and features in a virtual environment are finding they need to learn the skill set of the person who manages virtual machines.

4. Increased cloud deployments will put more focus on WAN and Internet performance.
The Internet is often the weakest link in a cloud deployment. In 2014, more enterprises are expected to shift their focuses and investments from private cloud services to public or hybrid cloud deployments. This means more data and applications get further away from users, who will continue to gain access from remote offices over the Internet. This will put added pressure on the IT department to ensure both the network and applications are fully optimized for a seamless and uninterrupted cloud experience. And since the Internet poses network bandwidth and quality issues, we will see WAN optimization deployment models evolve to address both SaaS and IaaS services.

5. Cloud-based replication and backup services will become mainstream.
As storage becomes more virtualized and cloud services take off, the cloud becomes a significant driver of network bandwidth demands. The business case for cloud replication becomes favorable, and as the cost of cloud storage keeps coming down, we will see storage administrators start to focus more on the total cost of ownership. 

6. Forced network hardware refreshes will accelerate the shift to software services.
Many customers will be facing a vendor-imposed hardware refresh in 2014. This will drive IT organizations to evaluate software-based alternatives. New flexible billing models and usage-based pricing will accelerate a move from hardware to software, and single purpose, proprietary Layers 4-7 hardware will start to become obsolete. 

7. Internet VPNs will displace MPLS networks and will eat into MPLS market share.
With cloud services becoming easily accessible over the Internet, users will start to realize that if they can connect to the cloud over the Internet, they can also use it everywhere else. Therefore, the managed service provider market will reshape itself around the Internet value-add. 

David Hughes is CEO of Silver Peak, a software-based WAN optimization company. 


What about the applications?

I think you allude to a couple of things that really resonate. There will certainly be more orchestration of workloads across networking, compute, and storage. I do think that means we will see the beginning of a new skill set. I don't think this spells the end of network engineering as I look at this as additive more than substitutive. 

But I was a little surprised to not see more about applications and abstraction. Most of the SDN promises end up being somewhat dependent on vendor-agnostic workload abstractions. I expect to see a lot more attention here next year. If heterogeneous solutions are ever to take off, this work has to happen soon.

I also expect to see a lot more work around analytics and data management. Automation will eventually require feedback loops. Having an automated trigger and a manual backend validation doesn't yield a ton of value. Data will be king. And there are tons of implications to that. For instance, do we see analytics make their way into controller platforms? If so, what does that mean for companies like Guavus and Splunk or any of the application performace monitoring companies like Boundary?

Should be interesting times for sure.

Mike Bushong (@mbushong)


Networking in danger?

David, great ideas here -- thanks for sharing them with us. Do you really think the networking profession will come "under attack" and in what way? Networking pros have been tasked with learning about virtualization, storage, cloud, for several years now. Do you think businesses will eliminate networking positions, or merge them with others, like the data center manager?

Re: Networking in danger?

Good question, Sue. My guess is that the position will morph into something else. Look what happened with server virtualization. Server sales are way down and my guess is that the old server teams are now virtualization architects. Am I right? 

SDN just one part...

SDN is not the answer, just one part of it. See SDN is NOT the next big thing. SDCI is.


Re: SDN just one part...

Dave, I beleive you are right, SDN is just part of a far bigger picture. But I'm not sure how many organizations have the wherewithal to embrace softward-defined everything (whatever you want to call it) quite yet. There is a lot of forward-thinking that may not be present on the business side of a lot of companies, never mind translating that to IT. I think it's definitely a great goal to strive for, but right now, different part of IT will still need to view it from their own reference points, be that networking, data center, storage, etc.

Re : 7 Ways Networking Will Evolve In 2014

@ Marilyn Cohodas, I don't think your server virtualization example applies to networking as a field. Networking is a complete discipline and not just about providing data services or security. Network as a physical entity, whether wired or wireless, will always exist and we would need people to design and maintain that.I can see added responsibilities for networking professionals, not the vanishing positions.

Re: Re : 7 Ways Networking Will Evolve In 2014


What if were to agree that in all likelihood, the role will change over time? There will be many companies who are not going to make the move to SDN any time soon, so their needs won't be changing. There will also be few who will be able to jump to a 100% SDN-based network either, so the need for traditional (legacy?!) networking will continue there too, and everywhere in between. However, in all likelihood most larger companies at least will be embracing automation and orchestration in the network and integrating that with the server provisioning and maintenance processes. This is a good thing in many ways, and means higher reliability when provisioning and migrating compute instances, but it also means less time spend by network engineers doing those jobs. Does that mean job loss? Probably not, but it does mean that people in those roles will surely have to do something else with their time - their role will change.


Re : 7 Ways Networking Will Evolve In 2014

I am a little perplexed here. How is virtualization separate from networking? VMWare or some other virtualization solution is frequently used by beginners at the learning stages because of the limited number of physical machines at home. So far as adding skills is concerned, that is true for all fields related to IT and not just networking.

MPLS vs Internet VPNs

7. Internet VPNs will displace MPLS networks and will eat into MPLS market share. 
With cloud services becoming easily accessible over the Internet, users will start to realize that if they can connect to the cloud over the Internet, they can also use it everywhere else. Therefore, the managed service provider market will reshape itself around the Internet value-add.


Interesting. I suppose it's possible that this could happen, but I'm curious about the idea of using Internet-based connectivity to replace a service that provides guaranteed bandwidth, often guaranteed latency, optional CoS, and SLAs. Compare that to an Internet-based VPN which, while cheaper, provides pretty much none of the above. Are you willing to stake your company's reputation on the ability to connect reliably to cloud resources via the Internet? I wonder how this will play out?