With many mature and proven open-source networking tools now available, many organisations have called time on buying the expensive products put out by big vendors. A brand new network built using free tools sounds great on the face of it, but what are the pitfalls? CNME investigates.
Networking has come on leaps and bounds in the past decade. And as the sphere continues to expand and diversify, encompassing both wired and wireless devices, all manner of commercially available tools have been released from the big vendors.
But while the likes of Cisco, HP and countless others promise much with their feature-rich tools for building up and managing networks, the open-source community has also stepped in with robust offerings.
Indeed, products such as Nagios Core 3.5, NetXMS 1.2.7 and OpenNMS 1.10.9 are now fully mature network monitoring solutions that promise users more flexibility than they’d otherwise get from big vendor products. Likewise, other tools have surfaced to help users build and define their networks, and they’re gaining steam quickly.
But is it right to go with open-source when it comes to networking? After all, if a company’s network goes wrong because it was set up using open-source technologies, it has no-one to blame but itself. What’s more, patches and fixes may not be as easy to manage as they are with the standard commercial products.
According to Den Sullivan, Head of Architectures for Emerging Markets, Cisco, however, it is possible to see why people would go down the route of open-source networking.
“I think that there’s often a perception in the market that the vendors — such as the Ciscos of the world, and other network technology vendors — have closed boxes. People think they’re closed and there’s a lack of innovation. I think, as a result of that perception that some have, they’re always looking for alternatives. We do that in our lives with consumer products, and we do that in our lives with professionals as well — if there is an alternative, I think we’ll definitely try to explore it,” he says.
However, he believes that people might be misguided when they cite a lack of innovation on the part of the vendors as a reason to move to open-source. “Cleary, from Cisco’s point of view, we invest over $5 billion a year in our R&D, which is more than 10 percent of the company’s annual revenue. That sort of innovation track is realistically not true,” he says.
That said, Sullivan admits that he does see open-source networking tools being used for the right reasons, too. He points out that network managers might use these tools in order to get a little more from their boxes, or else automate tasks. And from that point of view, open-source networking tools look reasonably attractive.
“What I see people use it for today is largely automation of tasks. And I don’t think that has really changed. That has been all the way through my career history. People have looked to try to do things faster, to automate things. And with regards to scripts and small programs, they’re taking open-source off the Web, bolting them together, and ultimately coming up with a little program or script that goes and does things a little bit faster for their own particular area,” he says.
Sullivan believes this is all that open-source networking is limited to in the region. And it’s easy to see why someone might be interested in exploring the option. Of course, another draw for exploring open-source networking tools is cost – many of these tools are available for free on the Web, meaning an enterprise could save a significant amount by not purchasing a commercially available tool.
That said, Sullivan says that the main reason why someone would explore open-source networking tools isn’t about saving money; he believes the main draw for it is trying to achieve more programmable flexibility. Indeed, he has his own ideas about the costs involved in open-source networking, and he says that the pitfalls to exploring open-source tools could very well negate the benefits.
“When you’re down there in the weeds, sticking it all together, building it yourself, when you can actually go out there and buy it, I think you’re probably increasing your cost base whilst you actually think that you may be getting something cheaper,” he says.
“Having been the CIO at Cisco for a very large region indeed, I think, really, that’s not the value of IT — to be creating your own thing, stitching it all together, whether it’s from various different vendors or whether it’s actually creating programs and actually building it all yourself. I think the role of IT is really around the issue within the business, and enabling your business to achieve its goals.”
According to Sullivan, building a network using open-source tools would demand that the head of IT was pretty much always on-hand to help with the work. This would mean that the CIO or IT manager would then have less time to work out how the IT department can best serve the business. In the long run, this could mean that building the network with open-source tools could lead to more money spent, as the IT head may miss important opportunities or problems.
“You look at where the business is going. Is that about you getting faster to market, is that about growth in the market, is that about going into new markets? What is that? How does the IT group align with that strategy, and then how best do they deliver it? Ultimately, I don’t think that is always about going and building it yourself, and stitching it all together,” he says.
“It’s almost like the application world. Say you’ve got 10,000 sales people — why would you go and build a sales tool to track their forecasting, to track their performance, to track your customer base? These things are readily available. They’re built by vendors who have got years and years of experience, so why are you going to start trying to grow your own?”
Perhaps, then, that’s why open-source tools aren’t widely used to build networks in the Middle East. Indeed, Sullivan claims he “doesn’t come across it all”, aside from when it’s limited to adding a little more flexibility to the network. That said, Sullivan does admit that the word “open” resonates with a lot of CIOs in the Middle East, but that instead they should be concentrating on software-defined networking (SDN) in order to achieve more openness.
“Of course, I see people very interested in the word ‘open’, in regards to software-defined networking, but I don’t see them actually going and creating their own networks through open-source, readily available programs out there on the Internet. I do see an interest in regards to openness, flexibility and more programmability — things like the Open Network Foundation and everything in regards to SDN,” he says.
Whether a CIO is ready to take on SDN is another subject entirely — particularly given the lack of standards and the emerging nature of the technology. So perhaps it’d be advisable to explore proven open-source technologies if you can’t find what you want from the major vendors. That said, Sullivan advises caution when taking the decision to design and build a network using open-source tools, even if the project is only small-scale.
“While you may consider that you’re putting it all together, or you’re creating something highly customised specific to your area, is that fundamentally your role in the group? Is that value creation for your company? Is that value creation for your board in the direction that the company is trying to achieve? How much are you going to spend on it? There’s always that classic example of build versus buy. You’ve got to make those decisions based upon your own interests of the time,” he says.
The open-source debate may be raging elsewhere in the world, but in the Middle East, there doesn’t seem to be much of an appetite for it. Some organisations may pursue these tools for specific projects or to add a little more flexibility, but in terms of the network at large, it would seem that people in the Middle East prefer buy over build.