Cloud could help fix Govt IT paradigm: Hodgkinson

news Independent analyst firm Ovum said this week that developing and maintaining ICT capabilities constitutes an ongoing challenge for government agencies, with one of the organisation’s Australian public sector specialists noting that the utilisation of cloud computing services could provide an edge in an “unsustainable game of ICT snakes and ladders being played by many government agencies”.

According to Hodgkinson, a research director with Ovum and a former deputy CIO of the Victorian Government, Ovum’s assessment of the pattern of audit reports in recent years showed government agencies were unable to sustainably advance strong ICT capabilities due to constraints in funding, resources and skills.

“Governments’ demands for ICT-enabled policy and service innovation are outstripping their capacity to fund the ICT capabilities of agencies. Mature, enterprise-grade cloud services provide a solution to this dilemma. They deliver a cloud innovation edge to agencies, enabling them to benefit from access to sustainable world-class ICT capabilities at a lower cost than would otherwise be possible,” Hodgkinson said in a statement this week.

Although local cloud computing circles have traditionally steered clear of the public sector as being too security conscious to be willing to host data in what has been seen as risky environments, recent initiatives by the government’s peak ICT strategy agency, the Australian Government Information Management Office (AGIMO) have prompted a shift.

In April 2011, AGIMO released the final version of a strategic directions paper on cloud computing. The paper pointed out the substantial benefits of shifting to the cloud model, and outlined a phased transition model for services to shift to different types of cloud offering specific benefits and adequate security. At the same time, the Defence Signals Directorate had published a checklist paper to enable departments and agencies to perform risk assessments to determine the feasibility of employing cloud computing services in their operations.

In October 2010, AGIMO had launched a major purchasing initiative, ‘datacentre as a service (DCaaS)’ to inform the way it purchases infrastructure-, software- and platform-as-a-service over the coming years.

Hodgkinson stated that a range of attributes characteristic of the cloud model constitute the cloud innovation edge, which helped cloud providers stand apart from the internal ICT capabilities of all but the biggest and most capable agencies. These attributes include focus, resilience, scale, multi-tenancy and iterative evolution, Internet-age security, use of service oriented architecture, self-service, social and mobile technologies and vendor ecosystems.

He said that agency ICT departments could rarely achieve or sustain all these attributes due to resource and skill constraints, and the challenges of supporting diverse and fragmented legacy infrastructure and applications. Cloud service providers score here, with their ability to define a catalogue of services optimised to run in a standardised infrastructure to world-class best-practice levels of performance.

According to Hodgkinson, the cloud innovation edge gives an important strategic perspective on the long-term benefits of the cloud computing model versus in-house ICT for selected applications and infrastructure services. Ovum predicts the maturity and sophistication of the leading cloud providers to grow rapidly and sustainably. “Unfortunately, the outlook for the maturity and sophistication of the in-house ICT capabilities of most agencies is a continuation of snakes and ladders development. Increasing fiscal austerity further compounds the likelihood that this will be the case,” he stated.

Hodgkinson proposed that agencies:

  • Include cloud services in their ICT strategy. A strategic perspective is required to position cloud services as part of a transformation of the agency’s approach to sourcing and managing ICT, he said.
  • Discover the cloud services available from trusted enterprise ICT vendors. Agencies may find that the most expedient way into the cloud is to leverage and evolve with existing trusted vendor relationships and procurement arrangements, he said.
  • Analyse application and data portfolios to identify cloud services opportunities. One of the advantages of cloud services is that they create both the imperative, and the opportunity, for agencies to focus on information and data rather than technology and software, he said.
  • Get hands-on experience with cloud services. Agencies should put selected applications and/or infrastructure services to the test and see the reality of cloud services for themselves, he said.

opinion/analysis
I believe Hodgkinson is right about much of this — Governments need to stop constantly trying to solve all of their IT problems themselves, or simply handing the issues off to an external outsourcer to solve. The cloud computing model can bring many of the strengths of both of these approaches together in a more flexible whole. However, I will note that cloud computing services in Australia are currently relatively immature, and that much of what Hodgkinson is talking about is not easy to achieve yet.

Much of the cloud computing infrastructure available in Australia is purely of the infrastructure as a service variety — many of the true cloud applications which would be useful to government don’t appear to have arrived yet; and if they have, I’m not sure to what degree they are mature yet. Personally, I would recommend governments look at targeted cloud investment — use it where it makes sense, but don’t try to force its use, as has been the case in the US Government to some degree.

Image credit: joegus74, royalty free. Opinion/analysis by Renai LeMay.

16 COMMENTS

  1. The biggest problem with cloud computing is that it is a buzzword. There is a lot of great potential in the technology and services driving “the cloud” but how can you expect the industry to mature when it has embraced such an immature name?

    • I agree. This is a tech forum, why use a buzz word for the masses? Call it the internet, that’s what it is. If your talking about offering processing serices on the internet call it cloud computing.

    • Cloud computing is a methodolgy that encompasses many areas lumped together. While the media may like to use it as a buzzward, it’s easier than having to state each of the terms seperately, IaaS, SaaS, virtual applications, virtual servers/workstations etc.

      The underlying architectures by cloud computing are very real, and hardly buzzwords.

      To suggest that the industry is using an “immature name” is incredibly short sighted, as it shows that you’re focussing on marketing material. There’s nothing wrong with the “cloud” moniker.

      • Yes, cloud computing terms are real. “The cloud” is the cloud computing term for “the internet”.

        Notice how many articals start syaing they are going to use “the cloud”? Well the government has used “the cloud” for years. If you mean “use cloud computing services” say so.

      • No one is disputing the term “cloud computing” only saying things like adopting “the cloud” as if “the cloud” is something different because you use cloud computing services.

  2. Hey, lighten up a little, guys, the name has come to mean a fairly specific set of parameters.

    After all, we don’t (now) see anything unusual or lightweight in talking about Googling, do we?

    • Everyone knows Googling is using a search engine. Does everyone know the “the cloud” just means the internet? I don’t think so. Some journalists seems to write as if they are two different things.

      • The “cloud” most certainly doesn’t mean the Internet. When you refer to using the cloud, you are referring to using something “aaS”. Whether that’s IaaS, PaaS or SaaS. People in the industry know what “cloud” means. If you don’t or if you think cloud == Internet, I’d suggest reading: csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf .

        Albert

        • “Bio

          Sam Johnston is Director, Cloud & IT Services at Equinix, where he provides cloud computing expertise internally and externally for Europe.

          Sam was previously the technical program manager for Google’s global tape backup program, where he worked in the Site Reliability Engineering team in Switzerland and gained deep knowledge of current, former and future Internet-scale server and storage platforms. Before joining Google, Sam worked at Citrix Systems and served as founder and CTO for a number of successful high-tech start-ups in Australia, France and Ireland. A computer scientist at heart, he was recently listed as one of the leaders of cloud computing, having been an active contributor to the cloud computing community since its inception (particularly in the areas of security, standards and interoperability).

          Sam has advised and architected pilots for various European enterprises (including one of the earliest and largest commercial cloud deployments to date, started in 2006 with 35,000 users) and presented at a number of international conferences on the subject.

          Sam earned his bachelor’s degree in computer science from the University of New South Wales in Sydney, Australia.”

        • “When you refer to using the cloud, you are referring to using something “aaS””

          Thankyou for using the term “the cloud” incorrectly. It is in fact the opposite of what you have stated. In a cloud computing sense “the cloud” id everything outside your cloud computing construct. I’ve been using cloud services since 2004, not many fancy acronyms then, they were just virtual machines. Rather than having physical servers and the pain of trying to have multiple instances running and trying to load balance them across multiple cores and physical machines virtualisation allowed all that to be invisible.

          Nice to see you have some exposure to programming by throwing in a ==. Good cut and paste skills from Wiki too. Cloud computing expert as you are trying to present yourself, seriously doubt it.

          • Wow. Angry much?

            “Rather than having physical servers and the pain of trying to have multiple instances running and trying to load balance them across multiple cores and physical machines virtualisation allowed all that to be invisible”.

            I think your superior knowledge has stumped me. How does just using “virtualisation” make load balancing “invisible”? In a IaaS, surely I’d need to spin up virtual instances, and spin up and configure a load balancer? That doesn’t sound very invisible to me. Or are you referring to PaaS? Terms such as though are fare more useful that discussing “cloud”.

            As for the issue of “the cloud” versus “cloud services”, perhaps your 2008 definition is correct. However, if someone said to me they were using “the cloud for their email”, I’d interpret that as they were using a SaaS for their company’s email, not that they were using the internet for their email, and get on with my life. It appears you have the time to correct them, which is great.

            Also, I’m not really sure how I’m “presenting” my self as anything. Or are you asking for a copy of my CV?

            Albert

          • “Wow. Angry much?”
            No, just sick of every Tom, Dick and Harry pretending he is an expert.

            “I think your superior knowledge has stumped me. How does just using “virtualisation” make load balancing “invisible”? In a IaaS, surely I’d need to spin up virtual instances, and spin up and configure a load balancer? That doesn’t sound very invisible to me. Or are you referring to PaaS? Terms such as though are fare more useful that discussing “cloud”.”
            I am no expert in cloud computing, but I have used the model. It was closest to the PaaS layer, though their terminology was different. Instances of the software were fired up on demand. Each would service 2-32 clients, the number of clients could change with time. With Microsoft it was a Windows platform, Sony used Unix. There are different configurations of cloud computing that your IaaS, PasS, SaaS model. They are just the current formal model in popular use.

            “However, if someone said to me they were using “the cloud for their email”, I’d interpret that as they were using a SaaS for their company’s email, not that they were using the internet for their email, and get on with my life. It appears you have the time to correct them, which is great.”
            Which is incorrect, but this isn’t someone just casually saying that they use the cloud for email. It’s lots of reporters who write IT articles aimed at IT professionals using the term. IT people who haven’t had exposure to cloud computing will probably start using the term “the cloud” to mean the opposite of what it means to people involved in cloud computing, people they they will need to talk to as cloud services are used more and more.

            “Also, I’m not really sure how I’m “presenting” my self as anything. Or are you asking for a copy of my CV?”
            Well, by people in the industry, I took it to mean you were claiming to be in the cloud computing industry. No I am not asking for your CV.

  3. AGIMO seems to be contradicting themselves with their deals. First govt departments are supposed to sign up for 10-15 years data center deal – now they’re told not to worry about physical infrastructure, just put it in the cloud.

    If any govt department has signed up to a 15 year data center deal, they’re going to be looking for something to do with it in 5 years.

  4. Hmmm … folks to stand back from all this a bit and look at the big picture. In terms of definitions, the NIST material is pretty authoritative: http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf.

    From a practical perspective, however, most folks still entertain the notion that ‘cloud computing’ refers to a continuum of ICT approaches which range from sexed up in-house ICT (i.e. ‘private cloud’) through to externally hosted somewhat dedicated ‘private cloud’ services through to the purest expressions of public cloud services such as Google Apps or Salesforce.com. IaaS, PaaS, SaaS etc. blah blah … this can be debated till the cows come home … but the debate is all pretty pointless unless you are a vendor trying to sell some h/w or s/w in the manner of a man selling shovels in a gold rush.

    My own view is that the only real way to think about cloud is AS A SERVICE … i.e. is it available today with the cloudy attributes in a manner that suits your business requirements? Cloudy is as cloudy does. QED.

    I actually no longer see any point in the debate around private/public. The only thing that matters from an enterprise perspective (government or private sector) is whether or not the service is functional and enterprise-grade. So, the big deal about cloud computing is that for the first time business executives can buy ICT applications – preferably via their ICT department – that already exist as a service to drive innovation … faster, better and cheaper. Kick the tyres, if it works and meets you needs, use it.

    I realize this view seems overly simplistic to many folks in corporate IT departments … what about this, what about that? etc. blah blah. Get over it. Stop trying to do things that you cannot do better, faster and cheaper than a mature enterprise-grade cloud service provider. Once you make this shift of logic you will have time to focus more on driving innovation through smarter use of information because you won’t be fighting fires all the time on commodity ICT issues that add little business value.

    The US Government’s Cloud First strategy is the right way to think. A mature enterprise-grade cloud service should be the first option considered for any new ICT application or major chunk of ICT workload, or when a big chunk of S/W or H/W needs to be renewed. Perpetuate investment in in-house ICT only when there is a specific senior executive approved reason to do so. This is the only sustainable way forward given the reality of constrained in-house ICT capabilities that most organizations face … IMHO.

Comments are closed.