According to Alan Priestley, director of strategic marketing EMEA at Intel, there is one very big barrier to adoption of cloud marketing: risk aversion.
That risk aversion is no doubt born of fear, if not of the unknown, then certainly of the sheer work required before organisations are able to benefit from the flexibility and, potentially, cost savings of moving applications to the cloud.
It isn’t just about the fact that they have built far-reaching IT infrastructures in-house over the course of many years. It’s also about the changes required in how the IT department is run and in the skills of the people who work in it.
Intel, though, has already shifted some of its IT to the cloud, as appropriate. “Intel uses cloud, both private and public, but we also have a lot of IT that we cannot and will not put into the cloud,” Priestley told attendees at this week’s Computing IT Leaders’ Forum, which focused on the management of hybrid clouds.
Feel the fear
However, when an organisation is already facing acute IT problems, the fear factor comes from not facing up to them. UCAS, the universities’ clearing service, has always faced a particular challenge: for a few days every year in August, demand for its services goes through the roof as students rush to secure a university place.
By 2011, the website through which everything had been automated was struggling to cope: a new approach that could handle the August spike in traffic was needed.
The solution proposed by James Munson, head of IT at UCAS, was radical: it would shift much of its computing services to the cloud. Not only that, but it insisted on a contract that would enable it to ramp up its compute capacity in August when it was all-action, and reduce it (and the price) for the rest of the year when the service is quieter.
“In 2011 and 2012, UCAS had problems being able to deliver the scale that was required for that intense period in the morning when everyone was getting their results at the same time,” said Munson.
“It was all hosted on-premise where we’re based in Cheltenham. We had created quite a complex infrastructure environment – some Microsoft .NET, some database, some Unix, different storage area networks all hosted there, and quite a lot of bandwidth that all needed to go in to that location on that one day, so it was not surprising that we were having ‘issues’.”
Furthermore, the architecture around which UCAS’s services had been built was monolithic, which meant changes required far-reaching testing and the systems lacked comprehensive monitoring. “So when things started to go wrong, we didn’t have great insight into what was going wrong and what was causing the problems. Something had to change.”
In late 2012, that change was decided: a transition to a public cloud infrastructure, with services shifting to a combination of Microsoft Azure, which made sense given UCAS’s existing .NET application investments, Amazon Web Services to host the organisation’s Oracle databases, and Rackspace, with whom UCAS already had a relationship.
The key services for finding and tracking courses were rebuilt in Microsoft Azure, with in-house Oracle databases upgraded, re-engineered and ported across to Amazon. These are load-balanced across two zones and the website is hosted by Rackspace.
Munson found that the new skills required of an IT department in the era of cloud are very different from those required to run IT in-house – a skills gap also found by Rocco Labellarte, CIO at the Royal Borough of Windsor and Maidenhead.
“We are looking at a whole new set of skills,” he said. “There are lots of kids out there with the right qualifications in terms of understanding the environments. But actually getting people with the practical skills who have ‘been there and done that’ is another matter.”
He continued: “We have broken down our skill sets into a three areas: one is to move to a monitoring team, which is effectively just sitting there, watching the large screens all the time and being able to react very quickly because we are maintaining the service integration element internally.
“The second is having commissioned technical architects that understand exactly how everything is put together, both from a hardware perspective, and from a networking, security and applications perspective.”
Finally, although the organisation may be outsourcing to cloud providers, there is still a need for technical architects that can inform the organisation how it should be done, on the one hand, while challenging providers and their recommendations on the other.
Going the extra mile to get appropriately skilled staff in-house can save a fortune, he added. “Having internal skills, if they are significant, provides a cost benefit. We have avoided about £500m in spend by having the right skills from the start,” said Labellarte.
Sometimes, the power of the mainframe is the most cost effective answer. Computing's Peter Gothard puts Computing's readers' questions on the future of the mainframe to IBM's Z13 expert Steven Dickens.
This Dummies white paper will help you better understand business process management (BPM)