There’s been a lot of speculation lately about VMware snatching up Engine Yard. Just last year VMware grabbed up Zimbra, SpringSource, RabbitMQ, and Gemstone Data Management (a SpringSource acquisition) and earlier this year announced a joint venture with Salesforce.com to create VMforce.com. This gives VMware an entry point into the platform as a service (PaaS) world by offering a Java development platform on top of Salesforce.com’s (force.com) delivery platform which in turn is a heavy user of VMware’s virtualization and cloud products. And from there one can surmise that they next step is to offer productivity applications (i.e., email…) to compete with the likes of Microsoft and Google. READ MORE »
Posts tagged PaaS
Last week I attended a panel discussion on cloud computing hosted by MIT Technology Review magazine and sponsored by IBM. The title for the session was Navigating the Cloud: Finding Solutions That Make Sense for Your Business. The panel, hosted by Jason Pontin, Editor in Chief and Publisher, Technology Review, included: Charles C. Burns, Vice President, Research, Saugatuck Technology; Miko Matsumura, Vice President and Chief Strategist, Software AG; Sean Poulley, Vice President, Online Collaboration Services, Lotus Software, IBM.
Going into this session I had set my expectation level at “medium” – that I would walk away with a few different perspectives, but nothing “earth shattering” that would dramatically change my perspective on the state of cloud computing. I wasn’t disappointed… The panel was a mix of “old school” IT and some “new age” thinking. Actually a pretty good mix, and I was pleasantly surprised to see the panel members occasionally get in each others’ face on a particular topic. But all-in-all I didn’t leave the session with anything new that made the think the cloud is ready for prime-time, mission critical enterprise applications. It is simply still too ill-defined, and this was reinforced by the panel members. All three more or less agreed on the point that cloud computing is an interesting model and holds the potential to shift our business thinking, but as of now it is still not the platform of choice for CIOs.
There were some interesting points and perspectives worth sharing, so bear with this potentially long post. My thoughts are in italics.
The cloud is about mass social movement and there is a “club of people who get it”. It is “social on the top” (attracts people) but is “anti-social underneath” (eliminate people). Once again, I’m not sure if I’m “in the club or out”… What is it about the cloud that makes it the domain of a few who understand it (re: my earlier post on this subject)? Maybe I am oversimplifying this, but cloud computing is not a major paradigm shift. Yes, it has lots of potential to reshape the IT economy, but at the end of the day it’s still about hardware, software and services – just packaged differently. I do think the perspective on top/bottom socialization is interesting – it does offer a ubiquitous platform that allows never before seen levels of innovation. And underneath the covers it’s about expense reduction – it can be a cheaper model in some situations.
The cloud allows you to “plug in” at different levels that best suit your business needs. These levels being BPO, SaaS, PaaS, IaaS and raw infrastructure. This is probably one of the best value propositions for the cloud and where I think it holds the greatest potential. I’m not sure I would agree that BPO is a cloud service, but the other levels provide you the flexibility to move functions to the most appropriate service platform – providing higher levels of service, scalability, lower cost – whatever your business driver may be.
The low cost of entry to the cloud is perfect for incubation. It allows lots of innovation with lots of failures, but with a relatively low cost of failure. “Experimentation on the cheap” as one panel member put it. I agree – especially at the SaaS and PaaS level. You can get in and out of the cloud fairly cheaply at these levels. A little more difficult at the IaaS level, and definitely harder at the raw infrastructure level where you have to invest time and money to build your own operating environments. But for building Facebook or force.com applications it is a very cost effective tool. And utilizing cloud-based email, calendaring and project management can be a cost effective solution for startups and small companies. The real value is that it shifts from a CAPEX to an OPEX model – placing the burden and risk of infrastructure ownership on the cloud/service provider.
The cloud is a better way of solving “old problems”, providing new services “not constrained by the firewall”. This was somewhat of a stretch for me, but I can get there if I try… Yes, it does eliminate some of the historical boundaries that exist between enterprises and provides a more fluid approach for creating extended enterprises, but I’m not sure you need a cloud to do this. It does address some of the balance sheet issues that CIOs have struggled with in the past and can have an impact on resource issues, but I’m not sure it’s the panacea to fixing the old ways of doing things. I think the economy will have to make a major recovery and allow investment across the entire IT infrastructure to solve some of the really nagging problems.
Myths about the cloud. In this section the panelists were asked what myths about the cloud they would like to “debunk”.
- Myth 1 – People will stop buying hardware and software because everything will move to the cloud. People will stop buying hardware and software because (1) the economy will not allow them to do so, (2) consolidation of the IT industry will continue to cause fear, uncertainty and doubt about from whom to buy and (3) virtualization will eventually fulfill its promise.
- Myth 2 – The cloud is a technological phenomenon. The panelist felt that it was a “social phenomenon” instead, which brought some raised eyebrows from the other panelists. But I tend to agree. I think it really is more about the social aspect of building a community of interest and innovation. But you have to look at how and where the cloud will be used. As I stated earlier, I don’t think the cloud is ready for mission critical applications – it’s about innovation and low risk experimentation coupled with non-mission critical application services (email, messaging, etc.).
- Myth 3 – The cloud saves money. This point brought the most response from the audience (laughter). Nobody, to my knowledge has built a financial model that shows that by moving applications into a cloud layer you actually saves money – at least from an OPEX perspective. What typically occurs is a shift of resources from managing the application internally to managing the application in the cloud. I will agree that there are examples of avoiding ancillary costs such as training where well-designed, intuitive SaaS-based services can eliminate this need. But what I have experienced is that you still need someone to configure, manage and troubleshoot the use of the application – regardless where it is hosted.
- Myth 4 – The cloud is easy to use. The panelists polled the audience to see how many attendees had actually moved an in-house enterprise application to the cloud – the response was “none”. This response negated the need for the next point which was that it is not easy to move applications between clouds. It is easy to create an application in the cloud, using the cloud providers development platform and standards. But once it’s in their cloud the likelihood of moving it to another platform is pretty low, causing the next manifestation of “vendor lock-in”. This did bring up an interesting discussion about the notion of cloud “intermediaries” – who provide services for designing, developing and placing applications in the most appropriate cloud layer – and then providing “transfer/translation” services to move the application to another layer of cloud provider. This may be the next new wave of hosting services.
Fears about the cloud. In this section the panelists were asked what they thought were the biggest fears about the cloud.
- Security (physical and logical) and the potential impact on the US economy – what if 200,000 small businesses running in the cloud are taken off-line by terrorists?
- Privacy – people do not know what they are giving up (one panelist referred to this as the “roach motel” syndrome).
- Separation of duties
- Once the infrastructure is autonomic the issues will move to the “human layer” – siloism will lead to tribalism which will lead to paralysis
These did not surprise me. They are pretty much the same ones you hear over and over. I think the biggest fear is paralysis – not being able or prepared to make a decision based on a “wait and see” attitude towards the cloud. I think now more than ever CIOs have to be strategic in thought and tactical in execution – creating small but impactful steps towards solving the IT crisis which we face. Waiting for the next hype wave to crest is very dangerous. CIOs today need to look beyond the next big thing that will solve their problem (client server, off-shoring, cloud computing, as examples) and get back down to basics…
Standards are still a ways out, but even when they get here it won’t make it any easier to move to the cloud. The consensus was that standards for the clouds will be just like all other standards – they will ultimately be driven by the consumer and even once they are in place, each cloud provider will provide “extensions” to the standards that will make it difficult to develop “cross-platform” or “cross-cloud” applications. This somewhat goes back to the discussion around cloud intermediaries whose primary value will be around helping enterprises sort through the morass of providers, platforms and standards.
Overall the session was worth the time and effort to drive up to San Mateo and spend a couple of hours listening to the panel. As I said earlier, nothing new or earth shattering. But it did help confirm my thoughts around the cloud and what’s top of mind with CIOs these days. They normally think about three things with respect to enterprise applications – Are they up? Are they safe? Are they cheap? Right now the cloud can’t answer those three questions. So it’s going to be a while longer (if ever) before we see mass movement of enterprise applications to this latest technological/social wave of IT. But in the meantime I see great hope for the cloud as a low cost incubation environment that will be adopted by lots of people and enterprises. So I’m still trying to figure out if I’m “in” or “out” of the club… stay tuned.