The Bloor Perspective: Linux in utility computing, Ellison the visionary and hotspot models

This week the analysts at Bloor Research consider Linux's role in major HP, IBM and Sun initiatives, Larry Ellison's software plan and public Wi-Fi business models...
Written by Bloor Research, Contributor

This week the analysts at Bloor Research consider Linux's role in major HP, IBM and Sun initiatives, Larry Ellison's software plan and public Wi-Fi business models...

HP, IBM, Sun and others are creating an imperative. Their infrastructure initiatives are all quite similar and aimed at the idea of virtualising the hardware layer. The driver for virtualising is that companies have been buying servers in an ad hoc manner, deploying on a one server per application basis. Consequently, they assembled server farms which have an average hardware utilisation of about 20 per cent. This is a waste of money and a management headache. So companies are very receptive to the idea of corporate computer resource that is both managed and efficient - which is what HP, IBM and Sun are talking about. However, if you talk the talk you are also going to have to walk the walk and, right now, what can be delivered doesn't amount to wall-to-wall virtualisation - or anything like it. So how is it ever going to be delivered? Blade technology, grid computing, automatic provisioning, SANs, NAS and so forth will play a part but for it to work, and work well, it will require a standard OS - and there's only one candidate - Linux. The easiest way to see the need for a standard OS is to consider why and how TCP/IP became a standard. It didn't happen because it was the best option or because it was purpose-designed to run a world-wide network with hundreds of millions of nodes. It happened because it was the only reasonable choice at the time. The same is now true of Linux as regards hardware virtualisation. It qualifies because it spans so many platforms. It also qualifies because it doesn't actually belong to anyone. It runs on most chips and is rapidly becoming the developer platform of choice. The idea is starting to emerge that you virtualise storage with SANs and NAS and you virtualise server hardware with Linux - thus making it feasible to switch applications from one server to another automatically and quickly. The momentum is building and Linux is set to become the standard OS for hardware virtualisation in large networks. Other OSes may eventually have to impersonate the characteristics of Linux or move aside. *Shake-up time* Never one to be shy and retiring, Larry Ellison is giving the impression that he's enjoying himself. Well-known for his abrasiveness and not being someone who suffers fools gladly, Ellison answered questions at a recent briefing for financial analysts, during which he cracked a string of jokes. Ellison appears to believe that the acquisition of PeopleSoft is only a matter of time. Pointing out that the largest enterprise applications vendor, SAP, is larger than the number two, three and four challengers combined, he doesn't appear unduly worried about any government blocking the deal. Asked what steps Oracle would take to push the deal forward, Ellison stated: "If the government says we can buy it - that's it." If allowed to continue pursuing the acquisition, the first step is to get the poison pill removed. And, if that doesn't work, Oracle will take control of the board at the next PeopleSoft shareholder meeting in June 2004 and push the deal through. Ellison is dedicated to the idea of an all-encompassing suite of applications, integrated into infrastructure technology. The reason is that integration, services and maintenance cost more than just purchasing a software licence - making best-of-breed point solutions too expensive and complicated. It is only when applications are tightly integrated and coupled to a good underlying technology base that companies can achieve the scalability, flexibility and reliability that they require. According to Ellison, there are four types of technology company: companies with products based on a single area of functionality; single product companies; best-of-breed vendors; and integrated suites. Only one type is going to survive. Of the first, Ellison points to some of the companies that have sprung up in recent years offering solutions for procurement and sourcing - or, in Ellison terms, 'one-feature' companies. According to him, most have ceased to be a factor and many more will fail. There are the one-product companies such as i2 in the supply chain market. Ellison doesn't have much time for them either, contending that integration costs are too prohibitive, making the software too expensive to run. Companies offering best-of-breed solutions may have more of a fighting chance but Ellison points out that implementing best-of-breed solutions is only a transitional strategy before moving to a suite when it becomes available. He points to Siebel, which he believes has a best-of-breed CRM suite - although it is expensive to implement since it lacks other applications. Looking into his crystal ball, Ellison states that there are three companies that are going to make it: SAP, Microsoft and, of course, Oracle. He doesn't even count IBM as a survivor in the applications space, pointing out that it has tried to compete in the market but its offering was flawed. According to Ellison, the shake-up has only just begun. *McHotspot? Maybe not...* How do hotspot locations make money out of wireless connectivity? Simple, by not selling it. The problem is that many businesses are jumping on the Wi-Fi bandwagon thinking they are heading for a goldmine. Many will fall off en route; many will get there and not find nuggets. The one vital piece of the plan is a financially solid business model. Simple calculations on the back of an envelope will do for a start. Let's do a Wi-Fi hotspot. Base costs are low, say £75 per month for broadband for half a dozen users and amortising hardware costs over three years. If you revenue share with a provider managing the service for you, you may get 50 per cent of the £5 per hour it bills the user. So that's 30 user hours per month to break even. Or with access prices dropping to sensible levels that's more like 30 user days per month. A user per day. Should be easy you think. But do you really want people camped with their laptops, occupying your expensive real estate for a few extra pounds? Depends on your business model. If it's geared to rapid throughput and making a margin from what you sell up to each person moving through, you'd rather they cleared off to let the next person in. So it's surprising that McDonalds should want to set up hotspots in 75 restaurants. Even more surprising, they plan to bill by the two-hour period. They hope it will drive traffic during off-hours but wouldn't it be better if they just sold more food? When your business model is measured in fast bites not bytes, perhaps your wireless hotspot access should be given free with the food or based on some kind of loyalty model. Better to give customers a reason to return than a reason to loiter. Of course the problem for most Wi-Fi users is they'll need to loiter long enough for their laptops to boot, connect and shutdown. However the much-predicted crash of the hotspot market by many negative analysts is as over eager as the positive hype by the bandwagon jumpers. In any gold rush, the only sure way to make a fortune is to sell shovels. Or at least recognise that gold miners are a new target audience for your existing business. Many would-be hotspot locations should recognise technology for what it is: An amplifier for existing business models. Just remember that too much volume causes distortion. Bloor Research is a leading independent analyst organisation in Europe. You can find out more at www.bloor-research.com or by emailing mail@bloor-research.com.
Editorial standards