As 2006 comes to a close, I can’t believe yet another year has passed. From a network analysis perspective, this year will be remembered for many reasons and form the catalyst for what’s to come in 2007. Wireless starts off the first three entries in my top 10 list. Disclaimer: the ordering below is not necessarily important.
#1: Solid Growth in the WiFi Market. WiFi will continue to experience healthy growth in 2007, more so in the corporate rather than public sector.
Private WiFi growth at the corporate level will be strong, especially with next generation WiFi (802.11n) being far more efficient and secure than current offerings. Plus, we’ll benefit from the pre-802.11n and MIMO experience gained from the consumer Guinea pigs. Thank you very much.
Thanks to corporations utilizing real next generation cellular broadband offerings rolled out in 2006, and once the cellular broadband providers convince consumers to pay an extra $600 per year ($50 per month) for their new broadband services, wireless hot spot growth will slow down. But on the other hand, I for one have already had it with paying for all those little extras on my cell bill, not to mention my cable bill. So maybe there’s still room for hot spot growth.
#2: 802.11 b/g is Not Dead! There were 25% more 802.11 b/g chips shipped in 2006 than 2005, assuring b/g growth for at least one more year as these chips find their way into devices. I expect more devices to operate at "g" rather than "b" speeds, going forward.
#3: 802.11n Settling Down. 2005 was a tough year for 802.11n when the standard suffered setbacks due to heated disagreements. Luckily in 2006, the Enhanced Wireless Consortium (EWC) took the bull by the horns, with virtually all vendors but Airgo joining forces. Since then, Airgo has been sold, EWC draft specs are published, and the WiFi alliance announced a new certification program for pre-802.11n products. Disclosure: WildPackets is an active member of the WiFi alliance.
#4: 10GBASE-T Passed. This is a big milestone to help accelerate growth of 10 Gigabit Ethernet. Up to now, we’ve only had the esoteric CX4 copper and great, but expensive fiber components.
People were laughing at 1 Gig to the desktop just a couple of years ago and now it can be found in near every NIC in 2006. Can 10 Gigabit be far behind? 10 Gig developments will accelerate in 2007 and be everywhere in 2008 – following the same ramp-up that 1 Gigabit has enjoyed over the past couple of years.
#5: Accelerated Data Analysis. Related to #4 above, Cisco has been instrumental in driving down the cost of 10 Gigabit Ethernet at the core. This increase in speed and sheer amount of data is growing much faster than the capabilities of our analysis devices, which includes anything that analyzes packets: packet shaping appliances, firewalls, IDSs, protocol analyzers, and so on.
Packet sampling via NetFlow on extremely high speed links is only a partial solution. PCI express to stream packets at 10 Gigabits into our collection devices is also not the ultimate answer to solving this problem. Even write speeds on drive heads are under scrutiny as a bottleneck despite parallel drives. Standard OS drive mounts to Storage Area Networks (SANs) solves the quantity of storage problem but not the speed. Furthermore, any processing by expert systems and IDS requires real CPU cycles. What’s the answer? Watch for some pretty creative approaches coming in 2007.
#6: Network Forensics Comes of Age. Call it the popularity of compliance auditing, retrospective analysis, lawful intercept, or even CSI, the buzz is catching on. Despite faster networks, more data, and the nature of our applications changing from bursty to steady with real-time streaming audio and video, flexibility will be the key in 2007. Only tools that give the user capture and analysis for large amounts of data in real-time, near real-time, or post-capture, depending on the depth of analysis and performance requirements, will be the winners.
#7: Growth in VoIP – with Caveats. Enterprises typically lease their private WAN infrastructure and do not use the Internet for critical business data and VoIP. Unfortunately, for economic reasons, many small to medium-sized companies continue to rely on the Internet as their geographically dispersed backbone. Meanwhile their traditional land-line and PBX systems are being displaced with VoIP on a daily basis.
Many kinds of competing appliances, tools, routing software, etc. get us QoS to the cloud, but then what? Worse, consider that Skype, a proprietary VoIP service over IP with eBay dough behind it, is going all out by rolling out the $29 per year unlimited calling plan. Linksys (Cisco) is even selling a Skype phone for WIFi. Internet attacks or congestion will become major business disruptors.
The house I grew up in had this wall-mount kitchen phone (yellow) with a rotary dial and I recall never losing a dial tone in some 18 years, storms and all. Furthermore, a call with static was quite rare and there was never a delay between domestic callers causing them to step all over each other. Cell phones are largely for convenience, namely mobility, so that we can talk in our cars and elevators. The tradeoff is quality and we must put up with clicks, silence, drops etc.
What will we trade off for disruptions in VoIP over the Internet? Certainly not mobility (as in talking while moving) since VoIP is largely static (or relatively so in a small WiFi area). Occasional disruption in the Internet here at the office and at home drive me nuts. Crappy cell calls drive me nuts. I still have my land lines.
#8: Aligning IT to Business Objectives. The old adage “I never saw a problem bandwidth couldn’t solve” need no longer apply. We can no longer afford to simply throw bandwidth at a problem, especially outside the LAN (inside the LAN, more raw bandwidth actually becomes much cheaper over time due to raw speed increases, economies of scale, and segmentation). So we begin to focus performance attention on our web servers, application servers, back-end data base servers, packet shaping appliances at the edge, and so forth.
Do our investments in such technology align with our business objectives? One way to measure this is the satisfaction level of our end-users. We need to literally think outside the box and focus on how our IT keeps our users happy and productive. Many companies offer ways to measure “end-to-end” performance, many of them proprietary and others, including WildPackets, embrace the open specification produced by the Apdex Alliance (www.apdex.org).
#9: Explosion in IPTV. This is a somewhat controversial area because it takes on many different forms from newscasts to iPod video to YouTube. The protocols are all different too (from streaming video to HTTP to file buffering and so on). Whatever it is, IPTV received a lot of attention in 2006 that will increase in 2007. The real question is how much of the big growth in IPTV over the Internet will hit corporate networks, and what will corporate America embrace or allow other than video conferencing? Certainly there is value in training and other uses. Be ready for new impacts on your network and possibly a new breed of worms/virus and other problems.
#10: Network Lockdown. 2006 was a milestone as more corporations than ever locked down their networks. Guest access is now totally taboo or limited to separate wireless-only portion walled off from the LAN, with limited outside access – such as HTTP only to get at content and email. HTTP email, ugh.
This is a natural outgrowth now that we better understand how to block hackers and attacks entering from the Internet. Next, expect further lockdown by enforcing policies/postures on desktops and laptops that extend beyond anti-virus installs. In 2007 there will be a major uptick in interest in Network Access Control (NAC). The only question is who will emerge from the pack: Cisco, Microsoft, or the Trusted Computing Group (TCG)?
Merry Christmas, Happy Holidays and all the best to you in the New Year!
