Cloud Asia ran from 30 May – 2 June 2011 in Singapore. It was co-organized by the Infocomm Development Authority of Singapore, who is investing heavily in Cloud to make Singapore a hub for the region, and Informa. In attendance were around 250 delegates from enterprises of all kinds, cloud technology suppliers, cloud service providers and telcos. The conference clearly demonstrated the convergence of IT and telecom. Attendance geographically came from the Pacific nations, throughout Asia, to the Middle East. It provided a ideal event to get the pulse of cloud computing across the region, understand the current vendor positioning and the yawning gap between vendor hype and the reality of many enterprise situations.
I ran a one day pre-conference workshop entitled Cloud Computing 101, below is a sample of a few of the slides. With 75 attendees at the workshop, with the same diversity as the conference, the discussions generated in the workshop were frank, insightful and entertaining. Its clear many enterprise have implemented virtualization and been able to achieve a couple of years of not buying new servers. Cloud being an evolution of virtualization is clearly of interest, but the confusion being created by vendors in cloud computing is stifling the market and frustrating most potential buyers. My parting message in the workshop was: Cloud Computing makes sense for some enterprises and some workloads (we discussed the criteria throughout the workshop), its implementation is the same as any IT project (cloud is just a technology and business model), so follow ITIL guidelines, identify a workload that makes sense for your specific business to which the business case is clear and select the most appropriate vendor for you and that workload (do not take any notice of all the fluff being pushed in the vendor hype.) And its OK to use multiple suppliers as you do not want to put your eggs in one basket given the vendor focus on lock-in, asymmetry of risk, and the confused nature of the market. For example, Cisco looks to be pulling back from VCE (Virtual Computing Environment) which has sort of become a monopoly in this space, with unfortunately monopoly pricing.
The conference kicked off with a excellent set of keynotes including Ronnie Tay (CEO of the IDA) who explained the generous programs and incentives the IDA has put in place to encourage adoption of cloud services by Singaporean businesses. Stanley Tan, the CEO of Global Yellow Pages, gave the most incisive presentation of the event. He’s an enterprise customer, and he told the conference cloud computing is being sold wrong. Focusing on the technology is missing the point, focus on the business value. Cloud Computing is like using an accountant, you don’t buy all the accountant’s time, just a bit of it. Of course this simplification highlights that Cloud Computing is little more than a cheaper version of hosted computing / services, which invalidates much of the hype being thrown at enterprises. Bill Chang the EVP of SingTel’s business group gave a presentation on their One Office Cloud Service.
Regulator involvement in Cloud Computing continues to perplex, as most outsourcing and off-shoring regulation has been completed made years ago, and that’s all their involvement should be in Cloud. Now the IDA’s focus on stimulating cloud computing adoption is a good success case, but its not regulating, rather encouraging through tax incentives and targeted investment. But there continues to be lots of rambling regulator discussion, this is another example of the ridiculous hype around cloud computing has resulted in a bunch of concepts being mushed together into an Orwellian view of the future to create something so nebulous regulators somehow think their involvement is required. Given virtualization and hosting achieved their broad success without regulator involvement its fair to ask why any regulation is required.
In the afternoon of the first day I gave a presentation on “Telecoms and the Cloud: Does it Make sense for the Customer?” shown below. Here I had some fun pointing out the hype with Cloud Computing is being positioned as a revolution, that word was used by several suppliers. We’re entering a situation were customers are being brow-beaten to a state of believe don’t think, just like IMS (IP Multimedia Subsystem), and we know what happened there. In the presentation was a fun analogy to the famous Burton Group “Death of SOA” slide. I reviewed the reasons to become a cloud provider and why CSPs have a role. However, I qualified many of the points, for example operators have a poor relationship with SMB (Small Medium Business) in IT. I’m a small business and treated as a consumer by my CSPs (Consumer Service Provider). The cost difference per MIP between how Google and many CSPs run their data centers can be high as a factor of 10-20. The end to end SLA (Service Level Agreement) isn’t available in most CSP offers, or if it is, there are geographic restrictions. In the work I’ve done talking to enterprises, in many markets, they have little confidence in CSPs being able to offer the best IT solution to meet their needs. Dell.com and BestBuy/Geek Squad are great examples of what it takes to meet the SMB segments’ needs in IT. I reviewed the cloud offers from the Global CSPs, who have IT focused divisions built-up over decades, or through aggressive IT acquisitions in the case of NTTdata. Moving into IT takes time, a dedicated division with IT not telco people is required.
On the second day there were some great presentations on the implementation issues of cloud computing from Charles Huh, VP Cloud Business Convergence at Korea Telecom. There were lots of discussion on OpenStack, with Rackspace and Equinix demonstrating how they’ve been able to move workloads between data centers and cloud providers. Such work is require to break the almost monopoly of supply from VCE, whose pricing is stifling the market. We also need the other suppliers (IBM, Oracle, HP, Dell) to step-up their game with Citrix and OpenStack to provide choice and better price competition.
At the conference there were many examples of how Cloud is being mis-used, for example: it was claimed mobile device management is because of cloud computing – it is not; cloud brokering is necessary to manage data across multiple cloud services, but isn’t that what SOA (Service Oriented Architecture) does? On the future of cloud, most visions assumed connectivity is ubiquitous so services and data are pervasive. This ignores the SMS I receive when I arrived into Singapore informing me that data costs $20 per MB, or the fact that I’m on a plane for 18+ hours without WiFi. Legacy infrastructure and applications will not go away anytime soon, some banks still have 30+ year old hardware and software running in their DCs (Data Centers). We’ve still got a long way to go on Cloud. Vendors need to stop with the BS, and focus on straight-forward value defined in business terms not TLAs (Three Letter Abbreviations) and silly technology concept mash-ups. Stanley Tan, the CEO of Global Yellow Pages, got it right at the start of the event, we’re selling Cloud Computing wrong.