Cloud Computing panel interview with Sun Microsystems at ‘Entrepreneur Country’

During the recent ‘Entrepreneur Country’ event hosted by Ariadne Capital I took part in a Cloud Computing panel interview session on behalf of Sun Microsystems, I was able to capture some notes after the session and have replicated them here.

The “Does every cloud have a silver lining?” session was hosted by Ric Francis of Ariadne and taking part alongside me were Charles Black, CEO and founder of Nasstar (a highly successful provider of ‘Desktop as a Service’ capabilities), and Kolvin Stone, Corporate Partner at Fox Williams (we were due to be joined by Joe Drumgoole, CEO, CTO and founder of Put Place, who I’d liked to have met, but sadly he couldn’t make it).

This article is a bit of ‘double A side’ with “Sun Microsystems, Cloud Computing and the premier European VC event ‘Entrepreneur Country’” where I describe the event itself. For further background material you may also want to check out my Cloud Computing overview article “Cloud Relationship Model“.

Here are the questions that we were asked:

  1. Surely we’ve heard all of this before in various forms and guises? What is different this time? Why will it work?
    …response #1
  2. For this stuff to become truly embedded it will need to move from the man in the street to the corporate. Corporate CIOs are a risk averse bunch especially when you move into some sectors (e.g. Financial Services). What will influence the CIOs’ buying decision?
    …response #2
  3. It was all very easy when you went out and bought or developed software, installed it yourself, ran it yourself, etc. Does working in the cloud bring new issues with regards to data ownership, IP rights, other legal issues, etc.?
    …response #3
  4. What is your vision for the future and where this goes?
    …response #4
  5. An audience driven Q and A session including responses to “What do you think of Microsoft’s Azure Cloud initiative?” and “What is Sun’s Cloud Computing strategy?”
    …response #5

I’ve divided the responses into roughly three sections; what I said at the time, what I captured from the other panel members, and any thoughts I have now reflecting on the responses we originally gave.

Surely we’ve heard all of this before in various forms and guises? What is different this time? Why will it work?

WTH: Yes we have, we have heard a lot of rhetoric in a similar vein for a long time now, promises have been made about similar initiatives in the past without the ability to deliver in a holistic manner. However four major reasons are creating a cadence around ‘Cloud Computing’ which indicate it is coming of age. Those are:

  1. Convergence – both Technology convergence and convergence ‘in the round’ bringing all these other elements together. A number of technologies have matured both individually and together to the point that this is technically feasible in a manner that is consumable by the ‘long tail’ and now the Enterprise.
  2. Economy – the current recession and downturn, coupled with the fact that this is technically achievable, demand that new models of accessing computing services are accessible to consumers and organisations.
  3. Familiarity – people are familiar and open to the idea of using services and infrastructure ‘on the cloud’.
  4. Openness – approaches to sharing, open standards, open source and ‘vendor neutral stacks’ are all key elements to why the technology is accessible and why adoption has become compelling and interest is being driven out.

CB: Felt that High Speed Broadband was key to the adoption of Cloud Services and that the lack of further investment in Broadband would inhibit take up.

KS: Re-iterated that maturing technology and convergence meant that the time was right for Cloud Computing.

Now: I see now that when I talk about Cloud Computing, I generally mean application level IaaS and PaaS, i.e. development and engineering platforms, whereas a lot of people are including SaaS into the cut. I’ve had a number of conversations where the person I’m speaking with sees little, if no, distinction between Google App Engine (a PaaS play) and HotMail (a SaaS play). High speed Broadband is important, but, for me, in the context of the critical national infrastructure of the UK and important for the economic viability of the UK. For Charles high speed broadband is critical for Nasstars service because their customers depend on the reliability, level of concurrency, and latency, of the network connectivity provided to access that service.

For this stuff to become truly embedded it will need to move from the man in the street to the corporate. Corporate CIOs are a risk averse bunch especially when you move into some sectors (e.g. Financial Services). What will influence the CIOs’ buying decision?

WTH: Two pressures, one internal and one external, I feel will primarily drive adoption, and that they are both analogous to previous ‘waves’ of innovation:

  1. Internal Demand; the PC revolution compared to the Cloud revolutionOne of the biggest drivers of the PC (and Windows) revolution had been the dichotomy, and apparent poor relations, between Business Departments and IT Departments within Enterprises. Prior to the PC revolution when a Enterprise’s Business Unit requested new electronic and computer enabled services there was a perception, real or imagined, that they would be given a straight “No” to their request, or at least a hefty price tag and possibly a long wait for the service to be enabled, plus there was always the added danger of the project not completing at all. Frustrations ran high amongst Enterprise Business departments when the advent of the Micro-computer lead to Business Department managers realising that for relatively little cost they could introduce computers onto people’s desks and use them to fulfil their own IT needs. For the first time users had access to technology without a central IT department dictating to them how they should use that technology. And so begin the proliferation of Desktop and Workgroup ‘server’ bound applications that as an industry we’ve only really rolled back into the Enterprise IT estate through rigorous integration and standardisation.The ease of Business departments in accessing cloud based computing facilities with a simple credit card transaction was very similar to the PC revolution when the growth in local applications first exploded across corporate. I see this as a significant internal pressure on the IT department and CIO driven by their own Business departments.
  2. Competitive Pressure; the Internet revolution and the Cloud revolutionWhen I was first involved with building out web infrastructure I recall that they were a large number of Enterprises that considered the Internet a ‘fad’, a temporary and passing fashion; who could blame them when even IT industry business leaders like Bill Gates didn’t understand what was going on with the Internet until as late as 1995. We all watched masses of corporates being over taken by the new pure play Internet companies, before elephantine they started to build their own web brochure ware and e-commerce sites. I expect this pattern to play out again and see small start-up organisations which embrace the cost saving and time to market advantages of Cloud Computing to ‘out compete’ many of the legacy businesses.

The progressive and savvy CIOs are beginning to understand these pressures, and some do already, they will be starting to do R&D; around adopting Cloud services and developing their Cloud Compute strategies. Early adopters will be looking to sign up cloud providers as partners to use by the business departments and, where possible, avoiding having to integrate services deployed across multiple dissimilar Cloud providers (with differing stacks and engagement models). Some will looking at providing internal cloud services to avoid many of these problems completely (and to be seen as providing future proofed and strategic capabilities to the business).

More reactionary CIOs will wait until they are forced to adopt the technology; sadly some of them even to the point at which it is too late and they are being driven out of business by the competition (as in the initial phases of the Internet revolution).

I don’t want to give the impression that the more cost conscious CIOs won’t be amongst the group leading the charge to adopt Cloud services, because they are genuinely going to be at the fore front of looking at how to get the best business value to IT spend ratios.

It was all very easy when you went out and bought or developed software, installed it yourself, ran it yourself, etc. Does working in the cloud bring new issues with regards to data ownership, IP rights, other legal issues, etc.?

WTH: I think this is very much an “undiscovered country”; I don’t think we’ll know much until we’re there. Legislation appears to me to be reactionary, and often set after the event.

What we know about doing business on the Internet will still remain as good experience when it comes to adopting Cloud models. Additional areas that the new Cloud adopter may want to consider are: privacy and privileged user access (who has access to your data? and what can they do with it?), regulatory compliance, data location, data segregation (and ‘sharding’), Recovery, Investigative Support, Long term viability (is your Cloud provider about to be bought out by your competitor?), competing with your Cloud provider (does part of your business compete with part of your Cloud providers?), Technical Implementation (what security technology do they have ‘baked’ in?), and Service levels (is the service available when you need or are legally responsible to be?).

Around privacy I have some specific concerns, for instance, laws governing the wholesale manipulation of large scale amounts of personally referable data is only really being worked out now. While certain companies promise to “do no evil” who is privy to how the information they are able to gather is used? And some of the largest Social Network sites in the World have still yet to find revenue generating business models, there are plenty of people who are uncomfortable not knowing what they might do to achieve profitability and what implication this will have legally. Similarly laws around copyright, distribution and ‘ownership’ are being put under significant pressure by the way people and organisations now have access to information and share data across the Internet.

CB and KS: Both of the other panellists believed that the Cloud model would make compliance easier; centralization of data away from a localized pc hard drive model will help the business manage compliance (please note both responses where much longer and complete than this, but I didn’t capture them as well as I’d have liked).

What is your vision for the future and where this goes?

WTH:There’s two parts to this; the evolution of Cloud Computing and the evolution of the Web itself.

Firstly around Cloud Computing there’s a great deal going on at the moment, including conversations around standardisation and integration of Cloud services. Early drivers include organisations like the Cloud Computing Interoperability Forum (CCIF) who are looking at interoperability across Cloud providers. For the most part I think were going to see standardisation around the virtual stack and platform offered to the developers much sooner than any actual standardisation between interfaces and meta-interfaces. That’s the advantage of the virtualised and vendor neutral stack; it’s inclusive, well known by the developer community and a ‘level playing field’. Increasingly we’re going to see services and capabilities built straight over the Cloud, with fewer and fewer infrastructure build-outs, certainly amongst the start-up community.

Unfortunately Cloud Computing hasn’t quite hit the peak of it’s hype curve yet and is certainly “No Silver Bullet” to peoples problems; however it is a fresh approach at how they can procure Service, Platform and Infrastructure capability and, to an extent, allowing them to reduce overall CapEx and offset some of that CapEx for OpEx.

Secondly around the way in which the Web is changing; I think that there is going to be two major technology revisions and one major technology improvement. One of the these technology revisions is that of combinative approaches to deliver functionality; real “Mash up”s that deliver genuine functional benefit, effectively the complex web stack. You could see this as a continuation of the Read Write Web and Web 2.0 but the multi-service enabled applications being greater than the sum of their parts, and although we’ve seen a little of this in Google Maps, etc., I think that it still due to deliver unexpected functionality. The next major technology revision is that of the semantic web, or rather more tagging, tagged data and micro-formats, will lead to an ever increasingly linked and related web, one the pundits are calling web 3.0; whilst this is interesting and will of course generate a variety of useful, innovative and supporting technologies it will actually act as a leap pad to another significant technology plateau. The major technology improvement is that of ubiquitous ‘agent based computing’; once the Internet is much more easily transposable, due to the semantic web and the wholesale tagging of unstructured data across it these agents will be able to ‘scuttle off’ and work for you in the background. At the moment search engines and their ilk use spiders and web crawlers to gather data, but imagine that Google’s web crawlers worked for you, coupled with improvements in analytic processing, doing tasks related to your workload and preceding it. For instance imagine the time when your ‘agent’ will access your diary, interpret your meeting and work needs, go off and collate data, distilling it down as research and presenting it for you in the appropriate format prior to your schedule; this will be the next great lead forwards enabled by computing and networking technology.

These are the major items, of course there’s loads more to come; from the evolution of Social Networks, the new approaches to “Web Science”, the coming explosion of analytic processing across the Web and Enterprise, how Enterprises will evolve due to the new architectural models driven from the web, how computing and computer science will combine with other disciplines and scientific branches to create exponential benefit, and, if it comes to it, the ‘Singularity’, a technology epiphany implied by a long-term pattern of accelerating change.

An audience driven Q and A session including responses to “What do you think of Microsoft’s Azure Cloud initiative?” and “What is Sun’s Cloud Computing strategy?”

Audience: What do you think of Microsoft’s Cloud Strategy?

WTH: Personally I think it’s an interesting, and yet difficult (for MS) play. There are two major issues that MS have to address; firstly public perception of MS as a Monopoly and someone they don’t feel comfortable sharing information with, never mind their corporate data on the ‘Cloud’, and secondly the apparent complexity of the Azure architecture. The early slides I saw showed a four layer architecture, but with each layer having four layers, so sixteen layers before you even look at how you deploy your services – ouch. I can see why they’ve done some of this, basically to make it easy to adopt existing .Net Client-Server applications in Azure and to provide a whole suit of capabilities, but this assumes that the majority of early Azure developers are going to be ones with existing applications and that Cloud developers don’t care about vendor neutral stacks, something I believe they do.

Audience: What is Sun’s Cloud Computing strategy?

WTH: Frankly I’m not party to the conversations around setting our Cloud Strategy, although I have been giving feedback from the field organisation and influencing where possible; my remit is the UK and Ireland and how Sun addresses that marketplace. However I think that our Cloud strategy will map to and exceed customer expectations. Customers in the UK are asking for three things when it comes to Cloud Computing and I expect Sun to have a way to engage at those three points of opportunity. They are:

  1. Cloud “web” services; i.e. people are asking if we have an online Cloud service that they can use, like Amazon Web Services and Google App Engine. Lew Tucker, CTO of Sun’s Cloud Computing business unit, has already confirmed that Sun is building an online ‘pay by the sip’ Cloud; and Dave Douglas, EVP for Cloud Computing at Sun, is due to deliver a major status update at CommunityOne.
  2. Cloud build outs; i.e. people are asking if we can help build Clouds with them, whether this is internal or external, public or private, or hybrids of these. We can help them with Professional Services to build these Clouds, Managed Services to operate them, and technologies at all layers of the Cloud stack, such as infrastructure, hardware, operating systems, databases, open storage, platform, applications, virtualisation, systems management, and automated provisioning.
  3. Cloud technology innovation adoption; i.e. I hear plenty of customers say that they are not ready to use external clouds, nor build one themselves, however these customer are saying that they’d really like to use certain components of our Cloud technologies. Virtualisation and automated provisioning are two of the most frequently spoken about, along with large scale and highly available MySQL installations, and again we can help them with a full range of Services and Technologies that meet these requirements. I expect that there will be increasing demand for the Q-Layer technologies in this space as people look at virtualising the ‘vertical slice’ (the complete set of components needed to provide an end to end transaction).