In May, Clutch released a report analyzing cloud computing education at universities and colleges in the United States. Through interviews with experienced professors and lecturers, we found that cloud computing lags in the US for a number of reasons, including cost of courses, lack of experienced staff, and the rapid pace of innovation in the industry.
What is the situation with cloud computing education like around the world, though?
In Clutch's research for its first cloud computing education report, online searches revealed a larger number of structured options available for studying cloud computing in the United Kingdom, as opposed to in the US.
Clutch interviewed two professors from universities in the UK to explore this trend and gain a greater international perspective. Dr. Thomas Erlebach of the University of Leicester and Dr. Raj Ranjan of Newcastle University share their experiences teaching cloud computing and explain how cloud computing education differs in the UK.
Dr. Erlebach and Dr. Ranjan discuss two factors that may contribute to a greater presence of cloud computing education in the UK, as opposed to in the US:
- More funding for research
- More flexibility among graduate programs
Please introduce yourself and your background.
Dr. Ranjan: My name is Raj Ranjan, and I am a Reader [Associate Professor] in computing science at Newcastle University. I came to the school in September 2015 and previously worked as a big-data and cloud computing researcher in Australia.
Dr. Erlebach: I grew up near Munich in Germany. I studied Computer Science and Mathematics at Technical University Munich and did a PhD in algorithms for optical networks at the same university. After that, I took up a post at ETH Zurich in Switzerland as an assistant professor in theory of communication networks.
In 2004, I joined the University of Leicester as a Reader and was promoted to Professor in 2007. My main research interests are in algorithmic aspects of communication networks, approximation algorithms, graph algorithms, and online algorithms. I mainly teach courses at the Masters level on topics such as algorithms for bioinformatics, networking, and cloud computing.
When was the MSc program in Cloud Computing introduced at the University of Leicester/Newcastle University? How long have you been teaching in the program?
Dr. Ranjan: The cloud computing module and accompanying degree started around 2013 or 2012 and both were driven by industry demand.
Cloud computing and IoT were very important for enterprises at the time. Once the government put out those priority areas, universities tried to shift the curriculum based on them.
Since its inception, the curriculum hasn’t changed; we’ve been using the same modules for the last five years but have undertaken a review of the degrees, and, from 2019, we plan to orient the course more toward IoT and big data.
Dr. Erlebach: Around 2012, we realized that cloud computing was getting bigger and there was demand from industries for graduates with cloud computing knowledge, so a cloud computing MSc program would be attractive to students.
We developed our MSc program in cloud computing during 2012/13, and the first cohort of students started the program in October 2013.
I’ve been teaching the course “Internet and Cloud Computing” every year since autumn 2013. My course covers networking and cloud computing.
In the cloud computing part, I explain the basics of cloud computing and the architecture of the data centers where it takes place, discuss issues of security and fault-tolerance, and cover scalable, distributed computing using the MapReduce framework (a programming model that allows for massive scalability).
Students also complete programming assignments where they implement MapReduce computations on moderately large, real-world data sets using Hadoop.
Our report identified three main challenges when it comes to implementing a cloud computing course in US universities: a greater than average price tag, a fast pace of innovation, and a lack of available expertise.
Have you encountered any of these challenges teaching cloud computing? Are there other challenges?
Dr. Ranjan: In the US, cloud computing is perceived as a technology area, so less emphasis is put on hiring research staff. When it comes to the UK, Europe, and Australia, we don’t face these problems. Large amounts of money were invested in IoT and cloud computing research. We have a lot of qualified teaching staff and haven’t encountered a lack of expertise.
Because Amazon, Facebook, and Google lead the cloud industry in the US, all the real experts work in the industry, not in academia. This is likely a cause for this challenge, but we haven’t had these issues in Europe and the UK.
In terms of costs, normally, when we teach our module, we ask students to set up Microsoft Azure and Amazon Web Services accounts. This is a cost-effective approach because those providers give free access to students for a certain amount of time.
We’ve tried to design library assignments around those free hours from the providers. It’s not very comfortable, and we cannot have large-scale experiments as part of the teaching process.
We have done this in 2016, but, in 2017, I set up a private cloud in Newcastle and gave students time using that infrastructure for coding practice. It’s a more cost-effective approach both for the students and us. It gives them a much better learning experience.
The cloud technology itself is not changing that quickly. The things that are changing are those that build on cloud computing, like big-data, IoT, service-based systems, and machine-to-machine communication.
I think that cloud computing has remained relatively constant in terms of virtualization of data centers and energy efficiency, but the new paradigms for using cloud computing are changing.
This is the reason for wanting to review our curriculum to include topics like IoT and big-data programming and show how they all come together to create solutions for smart cities, traffic management, healthcare, and so on.
From this perspective, the change exists. We cannot teach the fundamentals of cloud computing without teaching new applications and paradigms that build on them.
Dr. Erlebach: The cost of giving students full access to a commercial provider’s cloud computing platform would be an issue.
Initially, I considered this option for my course, but I found that it might get too expensive and that it is not necessary.
For the MapReduce programming exercises in Hadoop, we provide a simple installation in our own lab so students can run their programs locally, which avoids extra costs.
For understanding the principles of MapReduce and practicing the development and execution of MapReduce programs, I think this is sufficient to give students a good learning experience.
The fast pace of innovation in cloud computing is an issue. Educators need to review course content every year to see if any materials need to be updated.
On the other hand, the course I teach focuses on the principles of scalable, distributed computing, rather than the latest technology or platform that is popular but may be overtaken by something else in a year’s time.
If students understand the MapReduce paradigm and have gained experience writing MapReduce programs in Java, I think they can transfer their knowledge and adapt to using tools such as pig, hive, and spark quite easily.
The fast pace of innovation in current cloud computing technologies does not mean that the course has to be completely redesigned every year. It means that educators frequently need to review course content and decide which new developments need to be reflected.
Lack of available expertise has not been an issue for us. Several of the academic staff in our department are very knowledgeable about current developments in web technologies and cloud computing technologies, and our department has good industry contacts.
I had to learn about cloud computing and MapReduce when I developed my course, but it was a topic that I was interested in, and I did not find it an obstacle.
There seems to be a larger number of specific degree programs for cloud computing available in the UK compared to in the US. Is this true? If so, why do you think so?
Dr. Ranjan: This is all linked to the large funding available in the UK and the EU.
In the UK, the research party for cloud computing is The Engineering and Physical Sciences Research Council (EPSRC). They announced the priority research and selected cloud and IoT to receive large investments of money over the next five years.
We have skill shortages in cloud computing, so we want to use this money to hire new PhD and post-graduate students and use them to train the other students.
If we want grant money, then we have to teach and make research in this area. If Newcastle gets this research fund, they already have a curriculum for cloud computing immersion, so we can better train the staff on the research we do.
Major innovation in this area in the UK has been due to how governments link funding to the strengths particular universities have in the area in terms of teaching and research.
In the US, it works a bit differently, since the National Science Foundation doesn’t fund in the same way, and it may be the reason for the discrepancy.
Dr. Erlebach: I was not aware of this difference. Maybe UK universities tend to offer a larger number of named MSc programs that share courses and differ mainly in the core modules that are compulsory for students.
Only a modest amount of effort is needed to include an MSc in Cloud Computing in a portfolio of named MSc programs. This could be a reason why more UK universities offer such a specific program. But that’s just a guess.
Do you think universities in the UK are doing enough for cloud computing education?
Dr. Ranjan: I don’t have a full view of what is happening in all UK universities. Newcastle, in particular, is the leading university in the UK for the cloud computing and big data areas.
We have received £30 million from the UK government to set up a big-data institute at Newcastle. It’s a bit ahead compared to other universities, given that the government is pumping a lot of money into our institution.
We have £5 million for our doctorate center for the next eight years and have a UK National Innovation Centre grant for the next 10 years.
This funding fits into our teaching program, and I’m pretty sure that there are other things we’d like to do as well, but the funding goes to specific places. Not everyone gets a share of the pie. Overall, UK education represents a good percentage of the GDP spending.
Dr. Erlebach: I think students need to get a solid education in computer science as a broader subject but with sufficient specific coverage of cloud computing topics.
The principles underlying cloud computing should be taught in courses, and students should be given practical exercises where they can apply those principles.
To get useful extra experience, students should be encouraged to explore commercial cloud computing platforms in their own time (use free trials, etc.), or to set up and configure their own small clusters on spare machines.
Our MSc students have the option of doing a 3-to-8 month industry placement as part of the MSc program. Many of the MSc Cloud Computing students do this placement with a company in the cloud computing sector, where they can gain a lot of very useful experience in specific technologies and the actual use of cloud computing in industry.
Finally, the students can explore different aspects of the topic in projects that they complete at the end of their MSc program.
Overall, this combination of courses, placement, and projects can equip students with a good understanding of the principles and technologies of cloud computing and with the experience of applying those principles to real-world problems.
About the Author
Riley Panko is a content developer and marketer at Clutch, a B2B research firm in the heart of Washington, DC. Her research focuses on the cloud. Reach out with questions, comments, or concerns at [email protected]
More Research on Cloud
- Cloud Education Lags at Universities: Four Professors' Perspectives
- The Cloud in 2017: Trends in Security
- How Businesses Use Cloud Computing: 2017 Survey