Managing Director of Nixu Software (Nixu Group), Juha Holkkola, Speaks to the Editor of TCP/IP World about DDI and IPv6
Brief Company Profile: Nixu Software leverages Nixu Group’s world-class expertise in secure software development and information security by developing fully virtualizable DNS, DHCP and IP address management (DDI) end-to-end solutions. The Nixu DDI solution has been optimized to deliver the best value in DDI industry through a combination of decreased TCO and OPEX, high levels of availability and security, and bespoke professional services.
Unlike traditional hardware-based appliances that require specific server platforms to run on, Nixu DDI software appliances can be run either as virtual machines leveraging existing virtual infrastructure from VMware and Citrix, or natively on industry-standard x86 machines. Thanks to technology partnerships with industry-leaders such as VMware, Citrix, Hewlett-Packard and IBM, Nixu can be deployed as a standalone DDI solution or as an integrated part of larger infrastructure deliveries.
1. How was your company formed and what problem were they trying to solve?
Nixu was founded in 1988 by Dr. Pekka Nikander nowadays working as Chief Scientist at Ericsson Research NomadicLab. In the late 80s, TCP/IP networking was still at its infancy: National Science Foundation Network (NSFNET) utilizing TCP/IP was launched in 1986 and Finland joined the network in 1988, the same year Nixu was founded. So in the very beginning, being one of the few TCP/IP specialists in Europe with an entrepreneurial twist, Pekka’s aim was to help educational organizations, network equipment manufacturers and service providers to get up to speed with TCP/IP networking in general.
As the Internet started gaining further momentum in the 90s, Nixu got involved with a number of high-profile cases. We designed and implemented Inet, the first commercial ISP platform in Finland, for Telecom Finland in 1992. Later on in the 90s, we helped SSH Communications Security and Tatu Ylönen, the developer of SSH, to develop parts of IPsec. We also won a contract to design and implement an Internet supernode to an entire country, based on referral from Telecom Finland.
In the DDI field, we developed and launched version 1.0 of Nixu NameSurfer in 1996, after Telecom Finland had outsourced their DNS management to us after the Inet ISP implementation. Yet our go-to-market strategy was based mostly on an OEM model for the next 10 years, so it wasn’t before 2006 that we seriously started looking at developing end-to-end solutions for the DDI space. The 10 years we focused on OEM approach was quite helpful in building up a solid foundation for our existing DDI business, but we thought the time was right for a major revamp of our business model.
Around 2006 or so, we realized that server virtualization is likely to change the computing landscape in the same way as the Internet did a decade ago. So rather than launching a yet another hardware-based DDI end-to-end solution, we decided to remove the commodity hardware platform from the equation and focus on productizing the software server stack from the OS up, distributed electronically as an ISO installation media that installs and hardens the entire server stack automatically in about 10 minutes.
The major benefit of our business model is that it is much more efficient than the traditional hardware-based model. First, it allows end-users to leverage their existing virtualization investment for higher ROI through consolidation. Second, the cost of the DDI investment itself is lower than what Infoblox or Bluecat are able to offer. That translates to significant savings without having to give up anything in terms of functionality, redundancy or security.
Granted, it did take us until late 2008 or so before we got all the pieces right, but it paid off because we now make more virtualized DDI deliveries than anyone else in our space. Our customers tell us we are light-years
ahead of our competition in virtualized DDI solutions because Nixu has been designed with the virtualization-ready attribute in mind. Simply copying a server footprint and pasting that into a virtual template really isn’t the ideal way to virtualize network services, yet this is what pretty much all our competitors do.
2. When were you founded and how many employees do you have?
Nixu was founded in 1988. We now have little over 100 employees. We now have little over 100 employees and expect revenues of a tad over 10 million euros in 2010 (roughly 13.5 million USD at current exchange rate). Being a privately held and income financed company, sustained growth coupled with solid profitability has always been important to us because it allows us to provide the continuity our customers expect. Most companies in the DDI space have business models that require venture capital to be sustained. I think this is a fact of which buyers should be mindful when making procurement decisions and selecting DDI partners.
3. Will IPv6 affect your business? If so, in what ways?
Some of our customers such as FUNet (Finnish University Network) and ES.net (Energy Sciences Network in the US) have been running Nixu in dual-stack production environments for more than 5 years now. Thanks to these early adopters, we have been watching very closely how the IPv6 space has been developing. And of course, this also allowed us to introduce support for emerging standards such as DHCPv6 immediately after the specifications were finalized in late 2009.
In my view, IPv6 seems to advance in waves:
1) The first wave we saw started about five years ago and was driven by various research organizations and the academia. In terms of revenues or profits, I don’t think this was ever a huge opportunity, but it was quite helpful in being able to get input from early adopters of IPv6.
2) In 2009 or so, we started seeing an increasing level of interest in IPv6 from mobile service providers running networks in developing markets. These operators are often planning to provide ISP services over their mobile networks. Since many developing countries do not have that much IPv4 space allocated to them yet the populations can be rather large, IPv6 is often the only way to ensure the scalability of the service. We expect this trend to amplify during the 2010s and also spread to Western Countries after the IPv4 space has been allocated and service providers have exhausted the IPv4 space available to them.
3) The enterprise customers we are working with are not investigating IPv6 very actively. This is because they usually have private IPv4 spaces behind NAT, so they really are not in the risk of running out of IPv4 addresses in the near future. So we what we really expect here is a gradual move towards dual-stack networks as the existing network equipment and infrastructure is being renewed. That is likely to take quite a long time, though, because there really is not a sense of urgency in this space.
On the general level, I think IPv6 will increase the demand for productized DDI solutions, because of the complexity of the IPv6 address syntax. While a skilled system administrator was capable of managing IPv4 DDI services using visual editors and home-made scripts, IPv6 will make that increasingly tedious and error-prone.
4.What do you think the next area of focus will be in the Network Infrastructure and Management industry?
From the DDI perspective, we think these will involve consolidation of core network services (e.g. DNS, DHCP, IPAM) to virtualized server environments; more efficient management of name and address resources; gradual deployment of DNSSEC and IPv6 both of which require more advanced tools than organizations currently use; and technologies such as IMS (IP Multimedia Subsystem) that will be used in converged networks and rely on DNS/ENUM.
5. More and more daily consumer and business products are getting their own IP. What vertical Industries and geographic markets do you feel will generate the most demand.
In the big scheme of things, I think that the service provisioning vertical in developing nations such as the BRIC countries (Brazil, Russia, India, China) will be generating quite a bit of demand over the next few years. The demand stems from the large infrastructure investments made in these markets right now.
From Nixu Software’s perspective, I think the most important discontinuity relates to virtualization, though. Once all has been said and done, I think most network services and applications will be consumed from two kinds of clouds: the public ones run by service providers and the private ones run by enterprises.
Time will tell if and when this vision will materialize. But in the meantime, we at Nixu Software will be putting our money where our mouth is.