Tuesday, February 12, 2013

Hosted vs. Cloud, Which one is better for you?

You can find many articles online talking about Cloud vs. Hosted and why and lots of articles talking about how the cloud is the future os IT infrastructure, all you need is an Internet connection. Well, I personally agree, but may be not today. In many emerging markets the cost of bandwidth is still really high for universities, and even for universities who can afford it Internet is not reliable enough to simply move everything to the cloud, things as simple as copper line theft which is not very unusual in Africa can bring the Internet connection down for Days, backup Wireless links could be very expensive and not even available.

So for schools and universities in Emerging Markets they need to take that decision for every part of their infrastructure. In this post I'm not going to argue that cloud is good, secure, private and reliable because it is all that, rather I will focus on other parameters that needs to be considered by the IT team.

Where will the user use the service ?

I think this is the most important question to ask. If we are talking about a payroll system or personell the user is likely to be on a desk in the university most of the time so it makes sense to host such service in house. If on the other side we are talking about an eLearning service, the user will most likely access it from home to do an assignment or access reading material or an online course, for me it makes more sense to put such service on the cloud.

Many universities use Moodle and host it on a server in the university Campus, most users complain it is slow as uplink of the most universities are much slower than their downlink, availability is an issue given that most network maintenance take place after school hour which is when students access that system.

Email is used inside and outside the campus , it's used 24x7 any where in the w

Scalability

I know that this sounds like a cliché but think about the volume of email for few thousand students or think about the increasing amount of video uploads and the increasing size of course material especially with more HD photos and material available online.

Most cloud providers are experts in managing scaling, storage can scale from few hundred megabytes to terabytes very easily, increasing the number of cores that run a specific service may be much easier.

Scaling of a hosted service can be a nightmare for the university IT team. If the service may need to scale, you don't want to start with a large unused infrastructure then go for the cloud

Traffic and load spikes

Some services are idle most of the year then a sharp spike in usage. A good example would be the system used to announce the result or the registration system. These systems are typically used in the beginning and end of the semester and sit idle for the next of year. As a student, it was always a question of when the registration system will go down and not if, seats in many popular courses were limited and I remember that we used to wake each other at 6:00 to start registering once the registration is open at 6:30 each one refreshing the page every couple of seconds to check if the registration is open

Now you can host such system on premise, design your hardware and network to handle thousands of requests per day or per hour based on the size of the university and probably pay a lot for it or you can build such system on a cloud service such as Google App engine and only pay when you use this system during the spikes and may even pay nothing if you don't exceed your free quota.

So if you have a service that has high usage in limited period of time it makes sense to put it in the cloud.


What else?

Well that's what I have now on my mind, but there are several other good questions that IT teams need to ask themselves, if you have any more ideas, please share in the comments

Wednesday, February 6, 2013

Why I'm doing this

ICT in Education has been one of hottest topics everywhere and Emerging Markets is no different. Although that was not my plan when I graduated but I spent over 10 years of my professional career working on ICT in Education and I loved it.

I had the opportunity to work on several interesting projects, Deploying eLearning solutions to public schools in Egypt with millions of hits and thousands of online virtual classrooms, 1:1 classrooms in a school in rural ares of Abuja in Nigeria, the Magalhães project in Portugal deploying Intel classmates to over 300 thousand school students and training thousand of teachers on 1:1 learning concepts and finally helping universities in the Middle-East and Africa deploy Google Apps.

I've often run across programs with huge budgets that fail, or projects that start with no long term plan just for some government official to have his photo  in the newspapers or brag about bringing ICT to schools.

This blog will be about my stories with ICT in education, what I think works and what should governments or universities consider when planning for ICT in Education. It is also a space to discuss concepts and trials and use cases in the classroom from very part of the world.

I'm simply happy I'm finally starting this, and I hope someone will find it useful.