One of the long term objectives of all the big IT systems vendors is to find the next `big thing’. Selling servers that are now off-the-shelf commodity items, even with infrastructure management software, is now a recipe for disastrously decreasing margins. That next big thing is likely to be the big `S’ word – services. This is why IBM’s recent Blue Cloud announcement is more interesting for the future it presages than what it offers now.

IBM’s laboratories are most certainly not alone in developing technologies that allow users to request packages of services in the form of applications and tools together with the systems resources needed to run them. HP, for one, has been demonstrating just such a capability at its research labs for more than two years. The company may well be kicking itself for not announcing something already – and conspiracy theorists might even be tempted to ponder any possible connection between HP’s Software Universe event next week (a regular focus for new announcements) and IBM’s announcement last week.

Be that as it may, IBM’s Blue Cloud looks at first sight an interesting attempt at providing users with new tools that offer users a far greater operational flexibility. This is where users can start to request packages of applications and resources that can be provisioned dynamically, either when requested (if sufficient resources are available of course) or at a pre-determined time, for a pre-determined time period. Such packages would, of course, come to be seen by users as `services’ called up when required. This would mean that one of the underlying concepts of SOA – getting users to stop thinking in terms of `running applications’ (with all that that entails) – would at last start to become reality.

It would not be unreasonable to suggest that Blue Cloud, as it currently exists as a BladeCenter chassis of Linux-based Blades, is little more than a prototyping tool. But the question for CIOs and IT managers is then – a prototype of what?

It would be easy to suggest that the `production’ system would be a whole datacentre of BladeCenter chassis’ providing a gloriously flexible SOA infrastructure just for one company – an `intra-structure’. But, even with the developments in IBM’s autonomic computing environment announced a couple of weeks ago – where the systems can be expected to care for themselves even more comprehensively – having to provide maintenance and support to such hugely complex datacentres scattered all over the globe would be, to put an unfine point on it, an economic pain in the butt.

However, if those datacentres were fewer, and bigger, and IBM-owned the economics of it all could look a lot brighter, and they would be moving in on the `inter-structure’. In addition, the company would be selling systems to itself to provide a service to end users rather than to end users themselves, or other service providers. This would be an obvious opportunity for the accountancy profession to wax creatively in the area of operating margins.

Let’s face it, when compared to the network `Cloud’, Blue Cloud is a mere wisp of steam from a dying cup of coffee. But for IBM, HP and one or two others, moving towards owning a large tranche of the network Cloud that is not already owned by Google - and providing the services that users will need - is an obvious and important goal.

Getting users experienced in the new ways of working in and through the Cloud – and in particular away from the current psychological barriers to SOA that present themselves in the concepts of `my data’ and `my applications’ – is arguably the most important prototype work that could now be undertaken. Once beyond that, users can start to stop caring about `systems’. Who cares what `system’, `application’ or `operating system’ is being used, for the cloud is the cloud is the cloud.

Is the benchmark for benchmarks that they are no longer worth attempting? The question arises because far less action or brouhaha has occurred in the world of the new TPC-E benchmark than might have been expected. Both Dell and IBM have recently joined Unisys in publishing first results but, strangely, neither has made much fuss about them.

Back in July Unisys became the first server vendor to announce results against the new tests. These set out to simulate the online transaction processing workload of a brokerage firm. This is aimed at giving a much more relevant impression of how a server and database combination will perform in the real world than the old TPC-C benchmark, which while not discredited, has ended up fairly tarnished by rumours and innuendoes about the ways databases and hardware could `tuned’ to improve results.

Being first at anything like a benchtest, of course, will always leave any company on a hiding to nothing. So for a while Unisys had nothing to compare itself against, so no one - even Unisys - had the remotest idea of the relevance of the results. The company has an enviable reputation in many areas, particularly in building infrastructures and systems geared to the specific needs of vertical market sectors. As `a consequence it has usually eschewed the publicity-generating game of `who can build the real screamer’? So could the silence and Dell and IBM about their results have any relationship to that? The answer, at least in part, is yes.

It transpires that building a screamer is what Unisys appears to have done, at least in these early days of TPC-E. Its ES7000/one server came in with a performance of 661 transactions per second (tpsE), running a Microsoft SQL Server 2005 Enterprise x64 Edition database. Dell currently lies second at 220 tpsE, while IBM is third at 170 tpsE. Both of these were running the SP2 version of the same database. This result is no doubt coloured by the fact that the Unisys Server used 16 dual-core processors, while Dell used four and IBM only two.

However, when it came to price/performance, Dell and Unisys reversed positions. Arguably a more important metric for many enterprise users, this showed Dell at $1,020 per tpsE and Unisys at $1,777 per tpsE. IBM came third at $1,898 per tpsE, which might certainly explain its backwardness in coming forward.

The upshot of all this? Well, I do wonder if the reluctance to crow about these things publicly may indicate that there are now some doubts amongst vendors as to what benchmark results actually prove. For example, it is possible to read the price/performance figures to show that Dell can make a 4 processor server that is faster and cheaper than IBM or that Unisys makes bigger, more expensive ones – not much new there, then.

The performance figures can be read to show that the performance, per processor, of SQL Server drops off the more processors that are added, but that is hardly a startling discovery. And without then factoring in the different operating systems each is using – Windows Server Datacenter Edition for Unisys, Server 2003 Enterprise Edition x64 SP1 for Dell and the SP2 Version of that for IBM, getting anything concrete out of the figures gets to be a complex task. Can we say for definite, for example, that ipso facto the SP2 version is slower than SP1? I’m not sure I’d stake my life on it.

For IT managers, maybe this just demonstrates that the old ways of gauging these important issues are still best – get the sales reps to take you out to dinner and have `Scepticism’ set to `High’.

At the back end of last year Sun Microsystems came up with Project Blackbox, a good marketing wheeze in the form of a `datacentre in a box’. The idea is interesting but, given the nature of the box in question, it seems to be a little too specialist for it to be widely successful. In practice, as is often the case, there is a clear division between an idea and the specifics of an implementation of it.

For example, the `datacentre in a box’ idea has emerged again, but this time in a much smaller form and aimed at the small to medium-sized business community. What has that got to do with enterprise infrastructures? Well, quite possibly a good deal, at least conceptually.

Let’s go back to Sun’s box to make the point. This is due to appear around now and is some box. Yes, it can hold 250 SunFire servers with up to 2 Petabytes of storage. But it comes in a 20 x 8 x 8 foot shipping container which, although Sun argues is an economic form factor, is still not cheap. Not only that, the logistics of physically locating such a beast and feeding it with power and network connections will, I suspect, prove to be as logistically problematic for most enterprises as trying to build a similar datacentre inside an office block.

They could perhaps find uses providing the onsite resources needed in major civil engineering works, stacked amongst the Portacabin offices, but there has to be a risk of someone coming along with a big crane, a low-loader and a several sets of wire cutters…..bye bye box. All in all, it seems to suffer from sledgehammer/nut syndrome when looked at in the cold light of day.

And there the idea may have stayed, except that an Indian company HCL Infosystems, has come up with a much smaller implementation of the idea, aimed at the SMB sector. It is based on x86-architecture servers running either Windows or Red Hat Enterprise Linux, and the company says it has partnerships set up with ISVs to provide typical applications such as ERP, CRM, mail management and the like. So far so OK, but the boxes can also be clustered, the company has come up with its own solution for this task.

Clustered boxed datacentres based on industry-standard technology give enterprises the potential to build infrastructures quickly and flexibly – and even keep them indoors. It is an approach that is full of compromises, but that is what makes the building blocks of so many workable solutions. For example, it could allow the creation of a semi-distributed datacentre, where boxes offer sufficient performance to provide physical departmental server consolidation, while being integral to a logical single datacentre via high bandwidth networking. There is a compromise between energy consumption, centralised management, logistical flexibility and the provision of end user services.

It could also make a very good `quick and dirty’ compromise for scaling out services in a packed and logistically entrapped datacentre.

I suspect we may see the likes of HP and IBM moving to market their SMB-targeted Blade chassis systems along similar lines over the coming months as they realise that many enterprises, no matter what they might like to do with their IT infrastructures, come to learn that quick and dirty compromises are likely to be their best pragmatic solution.