Are schools really starved for bandwidth?
I was struck by a recent Wall Street Journal story on how school Wi-fi systems often have crummy performance. The conclusion the author and those she interviews draw: We need more bandwidth.
Maybe they do, maybe they don't. One thing I feel confident in saying is that nearly all of those schools are using what bandwidth they have inefficiently. Intelligent management of the network could improve performance for everyone. By the same token, sometimes throwing bandwidth at a slow network doesn't even fix the problem.
I spoke with Patrick Hubbard, "Head Geek" (yes, really) and Sr. Product Marketing Manager at SolarWinds, which makes tools for this sort of problem. Though they have school systems as customers, Hubbard pointed out that there are things that many schools can do with just the equipment they have. Hubbard knows about this problem, as he is on his kids' school's board of trustees, as well as the technology and budget committees.
The key is traffic flow management. With all due respect to the average school admin, they are not likely to be experts at traffic flow management.
Schools have lots of problems with tech, starting with the obvious budget pressure. The admins are also likely to spend a good deal of their time in their cars driving from school to school. Bandwidth options may well be constrained, especially if the school is far away from the central office. On top of that, many schools have the same BYOD issues as private business, only much worse. Few schools have any kind of mobile device management system in place, fewer still to handle BYOD devices, let alone the expertise to manage those users intelligently.
Under the reasonable assumption that a large percentage of these schools are using Cisco hardware, they can do a lot just with the capabilities built into those products. I'm talking about real Cisco equipment, not Linksys (which, until last March, was owned by Cisco). Other makers of business networking equipment, like Juniper, may have analogous capabilities.
First the admin needs to do careful network flow mapping and monitoring, for which Cisco IOS NetFlow works great. This will tell you who is using what protocols and connecting to which destinations. The admin can then build (for instance) a "Top 10 YouTube users" list or find users of gaming sites or protocols (or even worse). It's not necessary to shame people publicly; reaching out to these users directly and asking for their help in improving network performance for everyone may be all that needs to be done.
Second, the admin should implement QoS (Quality of Service), specifically Cisco's Class Based QoS (CBQoS). This allows the admin to prioritize or deprioritize specific typed of traffic. If a particular connection is important, CBQoS can guarantee a specified percentage of the bandwidth to it. Cisco's tools can project the impact on traffic of various changes.
QoS is not a new thing by any stretch, and yet it's entirely possible that the school's admins, even their consultants, don't know how to do it. This is a shame, and any network admin who doesn't know it should want to.
Featured
Hubbard describes other tech changes many schools can make to save money, starting with anything that improves remote access to networks from the district IT center, in order to make admins more efficient and keep them out of their cars. Working hard at this could allow the district to get along with fewer admins, especially if those admins are better-experienced and know more about network management. Alternatively, it could justify continuing education for current admins. The lack of expertise in key technologies in many districts is a big problem.
In the absence of expertise on using a network efficiently, it's easy just to say "more bandwidth." It costs more so it must be better, right?
And, in fact, the school may really need more bandwidth. It may be that using the connection to the best of its capabilities won't be enough. But if you employ these techniques and you still need to spend more on bandwidth, then at least you will have documentation that you put a serious effort in to use what you have as best you could, including historical data from the monitoring to show what improvements you made and what problems remain. Having charts can only help at budget time.
Another good strategy is to make a pitch to the district staff, letting them know that you're working on the problem and that there are things they could do to help. If the educators know that IT is working for them, then they will be more likely to back IT when it needs resources.