Through the fog... Storage as a service

Storage can seem so boring but the message is clear - ignore the latest developments at your peril. Here, Quocirca analyst Jon Collins argues that, at its most critical, storage is all about software these days…

Most IT managers like to think about storage in the same way as Victorian society considered a child - at its best when doing what it is told and keeping quiet. Storage does serve a purpose but, unlike other parts of IT (such as many applications, which get the majority of the attention), its ability to add value is largely ignored. Want more space? Buy more disks. And if the file and data repositories really start playing up, you can always implement a storage area network (SAN), the last word in ensuring your storage is seen and not heard.

Perhaps it is because storage has always been seen as an infantile element of the enterprise architecture that it has taken so long for it to develop and flourish. Let's face it, it was a good 10 years ago that networks reached some level of maturity, and that was with no little coaxing, cajoling and co-ordination from the companies that gave birth to them in the first place. While networks still have a way to go, this is largely because they continue to move in new directions (wireless hot spots are a current example). The underlying technologies and standards are largely dealt with, which is a lot more than can be said for storage.

Without dwelling on it here, there is plenty of evidence to suggest that storage is in a similar position to that of networking a decade ago. There is no universally agreed management standard (reference Bluefin versus Widesky), debate continues about the impact of new protocols (such as Storage over IP), and no single company has yet managed to provide a single point of management for heterogeneous storage environments, which not only enables monitoring and configuration of devices, but also supports services at the level storage administrators need to support. Think of a cross-environment guarantee of business continuity.

Despite this reality check, faced with the frequent apathy of their user communities, storage vendors are slowly but surely cottoning on to what is required of them. The belt-tightening of budget holders has helped, causing vendor companies to look for more innovative ways of selling their products and services.

Companies known for the qualities of their hardware are increasingly looking to software to provide the value-add in the shape of monitoring tools, cross-platform snapshot capabilities and increasing integration with the applications themselves, for example backup solutions that work with a database to ensure data integrity without compromising performance. These things are coming, indeed they are available from several suppliers (CA and Veritas to name a couple), but they are not yet considered as standard.

So what is the Shangri-la of storage? Is it really any more than providing one big pot that can guarantee to retain and deliver data? First and perhaps most obviously, there is only so much time that disk space can be added to an IT environment before something gives. There continues to be a data growth requirement, which might be increasing at the same rate as storage costs fall but it is not in line with administrative capabilities.

Second, in many companies the drive to reduce IT costs has thus far balked at playing around with data. Human psychology will always resist the desire to remove different versions of files, or even the same ones for that matter, unless there is an absolute guarantee of access. How many sales staff will keep the latest product files on their laptops in the knowledge that synchronisation with the corporate database is slow at the best of times, when it works at all? How many will also keep the previous versions, in the old format, which they find easier to work with anyway?

On the corporate network, even if there was a temptation to have a clear out, the fear of legal action provides a perfectly legitimate excuse to leave things as they are. In other words, and third, there is a growing need for more intelligent ways of managing storage assets, to assure the availability and integrity of data without costing an arm and a leg to do so.

From considering storage software as an insurance policy in case of disk failure, it seems inevitable that end user organisations will be clamouring for more savvy solutions for storage management. It's only a matter of time.

While companies may hear their storage arrays whirr with the sound of inevitability, it is equally assumed that it is up to the storage vendors to work out what shape this wondrous world will take. And the answer, ladies and gentlemen, is one big pot that can guarantee to retain and deliver data. Well, almost.

The idea is to give the impression of one big pot, by hiding all the gubbins of technology behind a software layer that can deliver the necessary storage services, as and when they are required. This has been tried before, of course, not two years ago and before the floor dropped out of the IT market. The buzzword was 'virtualisation' – a fine word when applied to the concept of a technology-hiding software construct.

Yet very few companies or individuals could agree on exactly what virtualisation meant. For some, it was applied to the firmware that simplified the addressing of logical units within the storage hardware. For others, it was the software gateway to storage subsystems from any manufacturer.

A third group went for broke, saying it wasn't truly virtual unless any type of storage could be accessed and manipulated from, well, just about anywhere on the planet. The vendors argued (politely of course – this is storage, after all), the market was confused and a consensus was reached to stop using the v-word any more.

They were all wrong, of course. A true virtual layer, whether high or low-level, is only of any use when it starts to deliver the services required by its users. This doesn't just mean disk space allocation, which is largely a solved problem. Rather, it is when storage can be delivered across multiple sites as a heterogeneous, ultra-reliable, cost-effective, secure environment that serves the needs of relational databases, media streaming, data warehousing, transformation and analysis, all manageable from a single console in a fashion that integrates with other enterprise management facilities such as the platforms from CA, HP and Tivoli.

Given this kind of capability, conversations around return on investment might actually start to have a bit more meaning, as the cross-environment costs of the allocation and use of storage can be measured with clarity and with meaning.

Nice dream but how can it become reality, given the ongoing gentle bickering and continued slow progress towards usable standards? Maybe the answer lies not with the big name vendors – the EMCs, HPs, IBMs and StorageTeks - but with companies whose USP is to provide the glue between them.

It's not dissimilar to what happened during the 'golden age' of enterprise applications, when the inability of applications such as PeopleSoft, SAP and Siebel to talk to each other, or to talk to older packages, gave rise to the need for enterprise application integration (EAI) products such as Constellar. In today's storage market, there is a growing need for integrated storage hub products that can deliver real storage services that integrate with the needs of business applications.

It is still early days but such products are starting to be available from players like DataCore and FalconStor, companies that have ridden the virtualisation storm and emerged the other side. These are not management platforms, rather they enable multi-vendor storage to be delivered in an integrated fashion. The level of services that their products provide is not right up there, not yet, but they are a good start nonetheless.

With advances in hardware technologies continuing to push the costs of raw storage down, the delivery of storage as a service is a software problem. Perhaps it always has been but there should be no doubt that to meet their ROI requirements end user organisations should focus on the software, not the hardware.

**Quocirca is a leading, user-facing analyst house known for its focus on the 'big picture'. For a full summary of its activities see www.quocirca.com, or reach the company's founding directors by emailing quocirca@silicon.com.

Also in this series: Through the fog... Buying an application server Through the fog... Corporate content management Through the fog... Automated speech recognition Through the fog... Public Key Infrastructure Through the fog... Vendor-channel relationships Through the fog... What future photo messaging?

For Quocirca's 'What's the fuss about...?' series for silicon.com, see this page

And for their earlier 'Surviving the Recession' series, see this page.