History 2004: Missing Information Lifecycle Management Links (Editor’s Opinion)

Information Lifecycle Management (ILM) is more of a general concept, a process even, with very specific underlying marketing strategies, than a product.

Initially, the idea came from vendors, and was so successful that there isn’t a PowerPoint presentation or press release by a storage company that doesn’t allude to ILM.

Even companies such as Brocade Communications or CNT, which have never made anything more than FC switches or routers, have taken up the call.

Overall, the idea is sound and its definition almost unanimously agreed upon. It entails the company managing its data, in whatever form (files, document, audio-video content):
– from cradle (i.e. its creation)
– to grave (i.e. its destruction)
– by way of the various stages of its lifespan, that is, indexing, processing, search, sharing or exchanging by workflow all the storage, backup, archiving and deleting
– with access speed dependent on how critical data is to the company
– with conversation compliance that meets current legislative norms
– all this in the most automated fashion, at the lowest cost possible.

For ILM, the hardware is ready to go, but that’s not the most important question. It isn’t conceivable to store all enterprise data on high-performance disk drives, due to the excessive cost. One highly appealing aspect of ILM is a system for storing different files on media with varied per gigabyte costs, depending on the data criteria: fast and reliable primary disks (FC, SCSI units), for databases, for example, less costly secondary disks (ATA/SATA) for less frequently accessed data or for fast backup and restore, tape or optical units for all archiving.

In order to be able to store data on the most appropriate device, will we be obliged to pre-define in advance, for all data, files or even email, its critical weight from its inception, and throughout any potential updates? With what degree of rapidity should we be able to access the data later in its life, with such processing choices evolving or adjusting according to the data’s intended lifespan? After how much time has passed can the data be deleted? Who will decide? And since that decision won’t always be made by the person who created the data, or modified it, should it be collective, involving multiple departments within the company? And that’s where we find the missing link of ILM.

For CAD/CAM or accounting files, the prospects look good, since they often contain the data necessary for this kind of identification and determination.

But what if instead, we take a more prosaic example: who’s going to be willing to annotate each entering and exiting email, each Word, Excel or PowerPoint file, with the criteria necessary to define their processing over the data’s lifespan? Very quickly, the average user will get tired of defining the criticality of the data, eventually settling on ‘Maximum’ every time.

While it is possible to find somewhere on the market software that allows for implementing some policy rules, i.e. such-and-such a file that is not used for a defined period, such-and-such a document that dates before a given date, or files created by a specific collaborator or employer, which are automatically transferred to a secondary and eventually tertiary storage material.

These criteria, however, are not enough, and currently, there is no single software package that allows for the automatic creation of a perfect classification, nor can we say one will ever exist.

To give just one example of this enormous difficulty, look at antispamming and anti-virus utilities that still are not 100% effective in separating the wheat from the chaff.

And the problem is, we can’t settle for less than 100% success in ILM. It’s always the little file that’s missing or not locatable that is so important.

Furthermore, how do we make modifications if the data’s lifespan has changed over time? And lastly, what must be done, if after several years the application program is not there anymore to open an old file? We’re not alone in thinking along these ways.

When asked about IBM’s missing link, Robin Burke, VP research at Gartner cited software in policy management. And that’s the rub, unsurprisingly. In comparison, the first and greatest impediment to effective document management was never scanning the paper, or storing it, but rather the heavy, onerous task of manual indexing. The second block is that the technology requires a constant re-examination of the company’s structure from top to bottom, its decision – making procedures and the responsibility levels of all its employees.

ILM is first and foremost a new way of organizing work. The method will have to follow.

J.-J. Maleval

This article is an abstract of news published on issue 198 on July 2004 from the former paper version of Computer Data Storage Newsletter.