Enterprise content management software is said to drive more efficient storage utilisation. So are the two tech-nologies converging? Madan Sheina investigates.
Could it be coincidental that reshuffling a few letters of storage giant EMC comes up with ECM, or enterprise content management, a suite of software that helps companies collect, classify, organise, store and retrieve corporate content in any medium or format in an efficient manner? Or is it a sign of real convergence between the two technologies?
It is true that the ECM space is becoming influenced by the storage industry, and vice versa. Market activity such as EMC’s acquisition of Documentum in 2003, and more recently IBM’s purchase of FileNet, seems to reinforce that view. At the same time, ECM vendors continue to tighten up their technical and marketing relationships with other storage leaders like NetApp, Hewlett-Packard, Hitachi Data Systems and Sun Microsystems.
There is certainly a case to be made for convergence. ECM software increases the need for storage and archival systems and when combined, the two create a loop of cross-over sales opportunities for both vendors. That opportunistic blend of technologies treads into a space referred to as information lifecycle management (ILM), which recognises a basic fact of data; that as it ages it becomes less valuable and needs to be moved down from expensive, high-speed top-tier storage, through secondary mid-range tiers, and then to tape archives.
“We like to think of ILM as simply providing the right information to the right person at the right time,” says John Rollason, EMEA solutions marketing manager at storage company NetApp. Meanwhile, Karin Ondricek, senior marketing manager for content management and archiving at EMC, likes to position ECM as providing an 'intelligent data movement layer' for ILM that migrates content to different storage tiers as business rules and policies dictate. “Having tiered storage is great. But it has to be able to respond to business needs that tell you which piece of content to allocate to the right tier, and when.”
ECM clearly has an important role to play in ILM. A 2006 EMC-sponsored AIIM survey, which investigated the role of ECM in storage decisions, found that larger organisations are aggressively pursuing consolidation and rationalisation of their storage and content archiving strategies. That is good news for vendors from both camps as the grey area between storage and ECM provides lucrative sales opportunities.
Carl Greiner, senior VP of infrastructure software and IT services at UK-based analyst Ovum, thinks it is a sensible move by storage vendors to expand their software portfolio as their hardware revenue starts to thin out: “Storage vendors like EMC could not have continued on a hardware path for much longer because of commoditisation and margins that are going to hell. Therefore, it makes sense for them to move up to the next [application] layer, which is where real business problems are solved.”
Storage vendors tend to agree. “Nobody wakes up one morning and decides they want storage. There is always a reason, a set of applications and data that is critical to a business, that drives that decision,” says Rollason. “That is making for some interesting combinations, particularly in the areas of records retention and e-mail archiving, both driven by regulatory pressures, that have a direct input into your storage architecture.”
EMC’s purchases of Documentum and Legato, an e-mail archiving company, could be seen as leading the way to some sort of technology convergence to fulfil EMC’s ILM strategy. To create systems that move data down through tiers of storage as it ages, EMC needed to add content management, back-up management and e-mail archiving software to its portfolio. But not everyone is convinced, not least because storage vendors have had little to point to in terms of real physical integration.
“EMC still has a very muddled story and other storage vendors such as NetApp, Hitachi and HP do not really have much of a story at all,” says Alan Pelz-Sharpe, a principal at consultancy firm CMS Watch. “There is a relationship, but it is complex and not simply about product convergence.”
While EMC has positioned its ECM products under the umbrella of a single automated ILM system, there has been very little attempt to integrate them more closely with its storage hardware. Integration has been limited to convergence of Documentum and Legato so that the two products store data in a common repository.
Storage vendors acknowledge that integration is less about physical integration and more about alignment. David Caldeira, director of ECM at IBM, thinks storage and ECM are complementary and believes that hierarchical or tiered-storage strategies should be determined by how businesses use their content. “ILM tells us that valuable content needs to be placed on expensive storage while less valuable information is put on cheaper storage. To do that we need to understand the business context of that content.”
Caldeira explains how ECM can map content to storage devices using policy-based content management, founded upon the overall business value of the content to the organisation, which he says is derived from a number of factors such as type of content, age, associated business processes, content state, as well as compliance requirements. “It really boils down to matching that value up to the cost of the storage infrastructure... not some theoretical storage strategy.
As companies evolve towards more sophisticated uses of their content, using that contextual business logic to drive your storage strategy is key, according to Caldeira. “[Storage] hardware is a by-product. What is more important is to decide how important the content is to your business, how quickly you want to access it and where to store it at the most cost-effective price. That is a policy decision... not a hardware decision.”
For that reason, storage vendors need to tie in to ECM software that provides a higher level metadata model of content. “What is needed is greater communication between storage boxes and the software, such that policies and rules established in ECM applications get pushed down into storage logic,” says EMC’s Ondricek. That, she adds, comes from tapping into the metadata inherently captured as part of ECM processes: “Things like who created the document, which business applications use it, how often it needs to be accessed and whether it needs to be retained and securely encrypted for regulatory compliance purposes. All that business logic informs the storage device.”
Darren Boyton, solutions strategist at ECM vendor Hyland, calls this a '[fit-for] purpose-driven strategy' that focuses on the business use of information rather than technical considerations around data format and storage source that will have to be put into place.
Yet the fundamentals of ECM and storage seem to be at loggerheads with one another. If the goal of ECM software is about managing the minimum content that is useful to an organisation, with as little redundancy, waste or duplication as possible, storage, which looks to store as much data as possible, seems to go directly against the grain “A storage salesman’s goal is to sell you more storage,” says Pelz-Sharpe. “That is why storage deals are huge in comparison to ECM deals, which themselves are not cheap.” But other analysts beg to differ.
“If information content is growing at 50% to 75% per year, as some have suggested, then it is impossible to keep up with that,” says Greiner. “We cannot simply keep throwing more disk arrays at the problem. Rather, we have to make storage more efficient and make sure that it exploits the intelligent functions in the ECM application layer more fully.”
Ondricek thinks there are sufficient synergies with the storage efficiency that an ECM system provides through techniques such as file compression, allocating and routing content files to the right storage devices with appropriate service levels, and tiered-storage based on business value drivers. “Paper is thrown away but digital content is usually not. ECM can help with that retention policy and combat the endless growth.”
Stark storage economics are also important, according to Kirk Roberts, president of Open Text’s Livelink ECM division. “Storage is one of the biggest capital costs for IT. The question is how we can smarten it up to reduce that cost.” He believes that ECM logic can lower it quite dramatically. “Even a simple de-duplication policy that erases 90 repeated e-mail attachment files rolling across a company’s storage infrastructure can lower your costs considerably.”
Some do not agree that customers are matching up their ECM software to their storage hardware, pointing out that Documentum, FileNet and other ECM products are designed to be entirely storage-neutral, even though some are now owned by storage vendors. “If that is the case, then why on earth would you trust an ECM decision to a storage vendor?” asks IBM’s Caldeira. “If you are a pure storage vendor you do not want ECM taking terabytes out of your business.”
But Pelz-Sharpe says that “in a typical ECM stack, the layer that has the least lock-in is the storage layer”, and hence believes it is not uncommon for customers to demand that their ECM platform be storage agnostic, even if they are currently tied to a single platform. That is a point that vendors who sell both storage and ECM technologies seem to understand. “We are truly agnostic in both directions,” says Ondricek. “Our Documentum software runs on any hardware and our hardware can run any ECM software.”
IBM’s Caldeira notes that the buying behaviour of storage and ECM technologies remains independent and has not converged in the minds of buyers. “Because the storage buyer is not the ECM buyer, we are conscious of remaining as storage-agnostic as possible. That has always been IBM’s strategy even though we offer our own storage hardware. For example, we have over 100 FileNet P8 customers running on NetApp’s SnapLock Check protective storage system right now.”
He acknowledges that this might be good for customers, but bad news for storage vendors, but it is a bullet they have to bite. “Whatever ECM software you buy, it will run with whatever storage systems you already have. Of course, if you buy ECM from EMC, they will try to sell you EMC storage. Ditto for IBM. But if you told EMC that you wanted to use Documentum software to manage data stored on NetApp or IBM disk arrays, EMC is not going to say, ‘Sorry, no sale’.”
Some sceptics would also argue that EMC and IBM have to be hardware agnostic not by choice, but by necessity. “Because storage vendors have acquired ECM software it is in their interest to be hardware agnostic,” says Simon Baumber, a consultant at UK-based IT consultancy Morse. “Supposedly all components will work together. But in reality they are still an amalgamation of acquisitions and even on the software side vendors have not done much to integrate these components.” Because of that Baumber has yet to see an holistic, all-encompassing ILM offering that covers both hardware and software.
Once thought of as separate IT disciplines, the worlds of storage and ECM seem to be coming together in interesting ways. Escalating content certainly increases the need for robust and scalable storage systems. And storage hardware vendors are keen to elevate their status in the ILM cycle. But what is not happening is a product convergence that some vendors and analysts would have us believe; thus far, we have only seen limited integration in the areas of records management and e-mail archiving. Do not be surprised if another big storage vendor such as HP buys an ECM vendor soon. But do be surprised if they start to more clearly articulate how they might add value to the ECM management layers that sit above them. So far, no one has achieved that.