In the next several newsletters we are going to discuss two technologies, both of which are likely to exert a profound influence on storage and storage management during the next few years. First we will look at information lifecycle management; then, we will examine storage grids.
This time however, we will look at these concepts not as separate areas of technology research. It is time to consider them as two important and interrelated parts of what may well be the enterprise IT system of the future.
First, some justification for ILM.
In a very real sense, the planets (or stars, if you have a more galactic point of view) are now in alignment when it comes to supporting ILM. I mean by this that the following indicators show that the time is ripe for the values that ILM delivers:
Point one: Pent-up demand. Many large IT shops have been deferring major storage investment for several quarters now. The result is that many, rightly or wrongly, already see themselves as being pushed to their limits when it comes to providing for their upcoming capacity requirements.
Point two: Budgets are getting bigger, but guarded spending will still be the order of the day. IT investment, while apparently a bit looser this year, is still going to be closely watched and stringently monitored. It is therefore reasonable to expect that storage investments will whenever possible try to extract maximum value, and that the concept of "good enough" will often replace "best in class" for hardware purchases.
Point three: "Good enough" hardware, when used appropriately, really is good enough for enterprise IT. SATA drives have already achieved widespread acceptance on the floor of the enterprise IT shop in virtual tape libraries and for storing non-mission critical data. (Some vendors are also trying to position a low-cost Fibre Channel drive - FATA - as a value-based alternative to the high priced Fibre Channel drives currently in use. FATA has yet to make an appearance ion the marketplace, however.)
Point four: The opportunity for flexible storage is about to emerge. Serial-Attached SCSI (SAS) drives are interchangeable with SATA drives (in SAS cabinetry), and one can be swapped out for the other without restriction. Soon, a single array will be able to provide two or more tiers of storage.
Point five: Moving data is not quite as painful as it once was. Data migration tools, while still apparently not widely used, are available from several vendors and allow some level of policy-driven automation to be applied to the process. Automated data movement between the various tiers of storage is the most crucial element for ILM.
Point six: For larger enterprises, storage grid technology now seems to be a lot closer than we thought. These have the capability of providing real-time scalability in terms of both capacity and throughput. If this ability to scale - all the automated provisioning capability that necessarily must accompany it - lives up to its promise, storage grids are likely to play a key role in IT, and in the largest enterprise shops may well provide the most efficient implementation of a site's top-tier storage.
More on the role of grids later.