Saturday, February 12, 2011

The Rise of the Network

It is quite profound how over the past 10 years the importance imageof “the network” has increased.  It used to be that you could happily get a lot of work done or communicate with others, offline or disconnected.  In schools if the network was slow or didn’t work, the teacher had a backup plan.  Often the network wasn’t critical to a lesson the teacher designed.  Today things are rather different…

In our School District (Coquitlam, BC) our success in infusing the use of technology for learning, teaching, and administration has now hit imagethe wall so to speak.  The network is our Achilles Heel!  I wrote about this last year in reference to a visit to a Digital Immersion 9 classroom.  Enrolment was forecasted to decline for this innovative school program where all students were expected to bring, rent, or borrow a laptop to use in this class.  I also referred to a consultative process I initiated around Digital Tools and Social Responsibility to uncover some root challenges.  In other words, this challenge has been growing, rapidly, over the past year or so once we reached a critical mass of participation in the digital realm.  What to do?!??

Part of the answer comes in questioning everything (as I did in Why?) to make sure how we’re using the network is sound, and what new imageinitiatives we’re contemplating do consider the dependency and impact on the network (we’ve asked people to go on a “network diet” and to pause computer purchases).  I’ve come up with a three pronged approach where we need to make three investments: (1) Increase bandwidth, (2) acquire and implement network management tools and add staff to run them, and (3) develop and implement guidelines for being digitally responsible.  Our staff development team and I (more them than I)have made great progress on #3 and are now going out to our education partner groups to consult with them on the draft work.  We are also revising relevant policies/procedures to be inclusive of the digital environment.  For #1, I have presented some options to our Board and they are digesting these in the budget process.

For #1, I’ve recommended building a private fiber network.  This is an expensive undertaking but in my view future proofs the District and provides the best value for money to create a network with legs.  Interestingly though, there are those that believe cell wireless networking is the future.  When I heard this I said “huh?”.  The idea as I understand it is that devices (ipads, laptops, netbooks) should be purchased with 3G (or 4G) wireless capability and data plans.  I question the wisdom of this suggestion from a value for money and a pure physics perspective.  A simple calculate for a 30 student set of ipads yields 30 (ipads) x 12 (months) x $30 = $10,800.  Extrapolate to 6000 devices and you get $2.1M / year!  I doubt most families will feel inclined to pay this (would be an inequitable expectation) and why would a school system?  Physics would suggest radio frequencies (wireless) will never compete with light frequencies (fiber) so as demand for bandwidth grows, wireless will tap out.  My assumption is that physical schools with hundreds of, or a couple thousand students and staff will exist into the future so connecting them at high speed, cost effectively would seem to be a reasonable expectation – also cell wireless isn’t ideal for high density (# of people) use.  Would it be appropriate for students to be connected to the Internet with no involvement (protection) from the school system?  That wouldn’t be my understanding.  But, I digress…

This leads me to #2 dealing with network management tools specifically (I won’t write here about the need for more staff).  From an IT perspective this is a fascinating topic.  It is amazing how far these tools have come in capability.  I am aware that quite a number of other School Districts have invested to varying degrees, in good network management tools – we have not (for a variety of reasons and due to other District priorities).  But it is now time!

In our research so far, we’ve (we have a great, although small, team led by Brian Lehmann, in Coquitlam focused on security and networks) discovered that there are a few fundamental components required (besides a good firewall design which we already have):

  • Packet Shaping / Prioritization
  • Proxy and Cache
  • Web Security and Content Filtering
  • Acceleration and Optimization

Our District has 68 school and 3 district locations which all route network traffic to a central data centre for District provided and Internet accessible services and content.  Preliminary advice would suggest that we need to acquire and implement devices as follows:

  • Packet Shaper (at the District core)
    • discovers applications running to/from the Internet
    • we prioritize applications (eg, for something like bittorrent, give it a trickle of bandwidth so it essentially has no impact; give youtube X% if needed but no more, BCeSIS Y%, e-exams Z%, etc.)
  • Proxy and Cache (at the District core and each school / site)
    • cache websites and video streams so that multiple accesses do not go to the Internet or beyond the school (note we have a cache architecture now but it’s old and limited)
    • eg, viewing youtube, educational streaming services without this requires that every person goes direct to the source but with caching, the 1st person does, the rest get it locally – huge reduction in network use
  • Web Security and Content Filtering (at the District core and maybe at some schools / sites)
    • eliminate malware and malicious website content
    • eliminate botnet activities (big problem for us at times due to us allowing personally owned devices)
    • reduce inappropriate content (this is a sensitive one and would need a lot of consultation, thought, and care as we wouldn’t want to filter out anything that is appropriate)
  • Acceleration and Optimization (at the District core and each school / site)
    • speed up applications that behave poorly (inefficiently)
    • cache content, documents, etc. locally and only transmit changes back/forth
    • would have a huge benefit to accessing and using remote files, Sharepoint (my43) content, e-mail and attachments, etc.
    • could allow us to centralize more services, save money on servers / storage, and increase our response time to issues and problems (although our people are really quite good already but users of our services always want more and faster)

One of our challenges, my challenges, is to find an optimal balance between paying for private fiber vs great tools to manage the bandwidth we have.  I see it as a both-and solution where over time, we have fiber everywhere AND are squeezing every ounce of performance out that we can.  We need to ensure that teachers, students, and staff experience no delays in using modern digital learning and work tools and content!

*ADDED TO POST* I presented an update to the Board on February 22, 2011.  The Tri-Cities News reported out on this update here.

 

After researching network management tools and contemplating the cost for installing fiber everywhere over the next couple of years, we developed and recommended a new option for the Board.  This new option would have us build fiber to our core (larger) sites and implement high performing network management tools at all sites to ensure all schools and District sites receive significant performance improvement in the short and long term.

For any of you reading this that have gone down a similar road, I would really appreciate any advice, cautions, lessons learned, etc. that you could share.  If you’re comfortable sharing brand names of products you’ve implemented and pros / cons you’ve experienced, that’d be awesome too.  Thank-you in advance…