What does the future of cloud computing hold?

Earlier this summer our technology forum surveyed heads of IT in further and higher education to find out more about the state of cloud adoption. It was quite a mixed picture - whilst 45% of respondees were using cloud services for business applications such as payroll processing, 31% had no plans to move to cloud for these applications. It was clear from the survey that many people were discouraged from using public cloud providers like Amazon Web Services and Microsoft Azure because of a perception that costs would be difficult to anticipate and contain. 61% of respondents said that financial issues were a major concern for them. We thought it was timely to do some further work in this area, and have just released a report exploring the future and potential of cloud computing for our community.

[Image credit: CC-BY Flickr user Barta IV]

Lessons learned

We’ve seen a lot of interest in the Google Apps for Education and Microsoft Office365 collaboration suites, which are widely used for student email, but by less than half of institutions for staff. These services have a host of other features beyond email that are relevant for teaching and learning and research, such as collaborative document editing and file sharing. Not only are these services free to use for education institutions, but Microsoft and Google have recently made unlimited file storage available as part of their services.

Whilst some institutions have approached them principally as “email replacements”, we believe that these services have huge potential beyond this core use case. But it must be said that we have heard from some institutions that have saved significant sums of money through moving to the Google and Microsoft suites - for example, the University of Westminster estimates that it saved £1 million on infrastructure costs through moving to Google Apps.

Of course these apps are free, which creates its own momentum and dynamic. Cloud computing infrastructure and platform services such as Amazon Web Services most certainly aren’t free, but can be very cost effective if used thoughtfully. For example, for just under £200 a researcher could create a cloud supercomputer with 500 processor cores and 750GB of memory for a few hours to carry out a computation. However, leave that system up and running and the bill would come to around £10,000 a month. At Jisc we run a lot of our services on Amazon and Microsoft’s Azure cloud, having carefully calculated our running costs, and we estimate that this has saved us over £100,000 a year compared with refreshing our old in-house infrastructure.

We see a lot of interest from the research community in complementing in-house high performance computing facilities with cloud capacity, and the report highlights a few high impact examples from the UK and elsewhere. For example, Mount Sinai hospital in the US uses cloud services to process around 100 terabytes of data for the Cancer Genome Atlas project, and Philips HealthCare manages a cloud based database of 15 petabytes of patient data from 390 million medical imaging studies. These examples, plus the UK’s own adoption of cloud technologies for government departments and the houses of parliament, prompt us to question the suspicion with which some people regard cloud technology.

Support from Jisc to move to the cloud

At Jisc we have been working for some time to help institutions to pick up cloud technologies. We have done things like put dedicated peerings in with the likes of Microsoft, Google and Amazon. We have created an institutional portal onto Amazon Web Services, which provides a consolidated view of institutional spending, delegated budgets for departments and research groups, and supports volume discounts. We also brokered national deals for the Microsoft Office365 and Google Apps for Education products.

Next steps

When we speak with the community there are a few key themes that keep coming up again and again - people are nervous about the potential costs of “getting it wrong” in moving to cloud services, and about being locked in to any one vendor. If we can find a way to reduce the friction of switching cloud providers and getting people the best possible deal, a bit like the uSwitch website, then it seems there will be a lot of interest in this.

We also hear that there is a substantial amount of interest in using technologies like Docker containers to reduce the amount of duplication of effort that takes place right now - perhaps in the future we can just package up applications once and then share them across the sector?

However the flip side of all of this is that cloud gives us a whole new lexicon and new technologies that we need to properly understand if we are to effectively exploit them. This is a journey we are all on together, and perhaps we can find ways to more effectively share what we learn along the way. This could be analogous to the way that we are supporting institutions in building digital literacy and digital leadership.

We are floating a few ideas in this report that we have heard from our stakeholders are important to them. What we would like to do next is have an open dialogue with the community about blockers and enablers and the potential that you see for cloud. We’ll be monitoring the #jiscfutures hash tag, and holding a webinar on 12th October as an open forum on the future of cloud for the sector. We’ll then work with stakeholders, cloud experts and cloud providers to develop a vision for cloud that can feed into Jisc strategy. We’d love to hear from you!

You can read the future of cloud computing report online, and sign up to our webinar on 12 October.

No comments:

Post a Comment