My Five 2011 Predictions

What's going to happen in 2011? First, we'll continue to see some cool technology and use cases in cloud computing and unified communications, but I think adoption of both will be slower than expected. We will also start to see more about Fiber Channel over Ethernet (FCoE) standards compliance and interoperation, particularly with FCoE and multipath routing protocols like Transparent Interconnection of Lots of Links (TRILL) and Short Path Bridging, though I am not convinced that the networking industry will rally around either interpretation. I think virtual desktops are going to remain as popular as they are now--not very popular at all. For each prediction, I am going to make a measurable statement, and I'll check back at the end of 2011 to see how I fare.

Cloud Computing: 1.) Private cloud use will increase by 20 percent. 2.) Public cloud use will increase by 5 percent. In 2010 the hype around cloud computing reached a fevered pitch. Everything is cloud, nothing is cloud. Cloud is going to transform your business. Cloud is just IT done right. Everyone had a take. Fact is, cloud computing has real, tangible value, and many organizations are looking at some cloud scenario in their future. I think that they are first looking at internal private clouds, meaning clouds that they host and manage in their own data centers. There is still a great deal of concern about the reliability and security of public clouds, which will slow uptake, except perhaps in SaaS, which is more "proven" for critical applications. In fact, in the Informationweek Analytics Research: Private Clouds report   [subscription required], 79 percent of respondents said they are interested in private cloud in their own data center, compared with 36 percent using a public cloud service.

FCoE Predictions: 1.) Cisco and Brocade FCoE switches won't interoperate. 2) SAN fabric hardware and software will have to support one vendor, or both. Status quo reigns supreme. Cisco and Brocade are going their separate ways in implementing FCoE (specifically FC-BB-5), which means that nothing in how you purchase and deploy SAN is going to change. You are still going to have to make a decision to go with Cisco or Brocade. HP hasn't made its plans clear on FCoE, so until product ships, the company is an unknown. You will still have to purchase from qualified equipment lists. Vendors that make storage fabric equipment will have to support both. You will still pay through the nose for SAN.

The fight over multipath Ethernet is going to start getting interesting. Two competing protocols--the IETF's TRILL and the IEEE's 802.1aq Shortest Path Bridging--should be ratified in 2011. Both protocols do the same thing, but Cisco has committed to adopting TRILL. HP has given every indication it will adopt it, as have other vendors. Shortest Path Bridging just isn't getting the same love, but that doesn't mean it is out.

Virtual Desktop Infrastructure Predictions: 1.) Our year over year research [subscription required] will show that VDI adoption of those deploying, piloting or testing VDI won't change by more than 5 percent over 2010. 2.) Of those using, piloting or testing VDI, the percentage of organizations planning on using VDI for all users versus some users won't change by more than 2 percent. 3.) VMware will increase market share over Citrix by 10 percent. VDI has been around in some form or the other for over 10 years. It's a very mature, very stable technology. The benefits of easier management, better security controls and better user experience have always been evident. What has held back VDI, I think, is that the cost of the server and workstation hardware, licenses and OS licenses have been on par with or greater than that of stand-alone desktops. Without compelling cost savings in operations, there has been no reason to end the status quo. I don't think that changes in 2011.Unified Communications: 1.) The number of enterprises using UC won't change by more than 5 percent. 2.) The number that have no plans increases by 10 percent. Our own research shows  [subscription required] that only 30 percent of enterprises have adopted UC, while 39 percent have no plans.

The remaining 31 percent are planning to deploy in six to 24 months. UC has some fantastic features for staying in touch with employees and collaborating over long distances, but the returns are often too soft to quantify, and the capital, operational and education outlay is large. That doesn't change in 2011. Perhaps in the years ahead, as users get used to video anywhere, anytime on mobile platforms, we will start to see a real uptake in UC.

Fiber Channel over Token Ring Predictions: 1.) Market penetration will be 0 percent due to lack of products in the market. 2.) Wikipedia will still refuse to host an FCoTR entry (or no one will want to put it back up). 3.) Some people still won't get it.

The main issue with FCoTR is that Ethernet was designed to be a lossy protocol, and that is how it works best. To make it lossless, it required no fewer than four fundamental changes to how Ethernet works. Maybe three, since congestion notification is a hot debate. With Token Ring, you had guaranteed capacity and zero loss, and you didn't have to deal with that steenking ATM cell tax. It's a technology past its prime, but that doesn't make it any less useful. A geek can dream.