These days I have found that most trade shows are hopelessly unimpressive, but I must say that Network World IT Roadmap Chicago was very useful. It was very targeted, so it was very relevant. Most of the presentations were good, the crowd was sufficiently nerdy, and the vendors were handing out swag again.
The swag award has to go to Cisco for their laser/pen/stylus (you had me at laser) doohicky, although the netscout bottle opener is probably going to get the most use. There were plenty of t-shirt giveaways and an iPod Scratch and win that I scratched, but didn’t win.
A couple of interesting stats that I heard there. The 3 year TCO (Total Cost of Ownership) for a rack of equipment in a data center consists of 60% of infrastructure charges. By 2012 it is expected that that will go to 95%. This is largely due to the costs of the powering and cooling infrastructure needed to deal with high density installations. A Gartner study was quoted that said 50% of all existing data centers are not capable of supporting high density installations.
Another stat tat was thrown out is that it costs on average $20,000 a year to keep an employee in an office. For traders and those types that need to be in the high rent district the price goes up to more like $30,000 and $10,000 out in the sticks. The point was that it makes a lot of sense to start to look at telecommuting. With the prevalence of broadband, VPNs, Network Access Controls, etc. the opportunity is there.
The hot topics seemed to be NAC (Network Access Controls), performance/VoIP monitoring, security and acceleration. Monolith Software was pushing a software based acceleration package that relied on converting the tcp traffic into UDP and handling all of the error correction and flow control in their software to maximize throughput.
Uplogix (uplogix.com) was showing off a console device that has a hard drive built in so that it can serve as a TFTP server and store last good configs for devices. It also has a modem so that it can be used for dial-in or dial-out. It is capable of monitoring the devices, so it can look at things like processor utilization, and can pull a show tech when the device gets into trouble and store the information on its drive for later retrieval. They were saying that this offers you a view into what is going on when your device is unreachable.
Another device that looked interesting was Route Explorer from Packet Design (packetdesign.com) that allowed you to receive routing updates from your network to determine routing patterns and changes. Since it is receiving routing updates it can alert you to changes as well as show you historical changes. It has a slick display that can show you traffic flows to identify things like asymmetric routes and single points of failure. It also allowed you to analyze proposed changes to your topology and to see what the effects of those changes would be. A bit pricey but definitely interesting.
On the performance analytics from there was Klir (klir.com) that was showing an outsourced network application performance monitoring tool. A small footprint collector sits inside your network to bundle up the information, encrypt it and send it to the Klir servers. I’m not sure how that would work with high bandwidth networks or for companies that are restrictive about that type of information leaving their network, but it was an interesting package.