Americas

  • United States

Automated computing: onward, ever onward

Opinion
May 08, 20034 mins
Data Center

* Automated tools and on-demand computing

The analysis of Enterprise Management Associates’ survey on user expectations of automated management tools continues to provide interesting results.  If you work for a storage resource management vendor, here is the good news: respondents answered overwhelmingly that automated tools offered great potential for increasing the efficiency of storage management.

The bad news?  Most IT managers are consumed with accomplishing the mundane day-to-day tasks of their job, and are much more interested in tools that help with that than they are with being provided with gee-whiz features.  Partly as a result of this, the majority of managers typically forego policy-driven tools in favor of going with their long-standing friend, scripting.

Vendors should think seriously about the implications here.  Automated SRM is expected to account for much of the growth in the storage segment over the next decade.  After all, it offers the only way IT managers will be able to manage all the data that will keep piling up. Vendors that don’t address users’ needs are unlikely to seize much market share.

We should ask why so many managers stick with scripting, which is difficult to extend and a challenge to maintain.

Here’s a hint: Scripting works, and the managers know it.

Even if scripting doesn’t scale, it addresses managers’ needs by running many of the simple tasks that an IT staff deals with in the course of their daily duties.  How to sell management tools into this environment?  The lesson here would seem to have something to do with marketing aimed at solving the problems at hand, and leaving the more sophisticated issues to be addressed further downstream once a vendor has its foot in the door.  Provide a fix for the simple but time-consuming problems first.

Now to another, related, topic.

There was a time, I think starting in the 1930s, when IBM actually had a corporate song book.  I imagine row on row of men in blue suits, white shirts and ties, marching in lockstep and singing the corporate anthem (“Onward, ever onward!” – if you don’t believe me, check out the amazing WAV file and words available online at https://homepage.ntlworld.com/barryf/songbook.html, I promise you it’s worth the effort). 

By the early 1990s however, it seemed as if every group at IBM was marching to its own tune.  Song book or no song book, the company was losing hearts, minds and market share, and was clearly in trouble.

I spent the better part of last week with the senior management team of IBM’s storage and storage management groups.  They were briefing analysts on their new corporate thrust of “on demand” computing, an important part of which involves provisioning new IT assets on an as-needed basis.  Significantly, a key part of provisioning involves “deprovisioning” assets when they are no longer required, freeing them up for allocation to other business demands. 

Automation sits near the heart of this concept. 

Consider the implications of dynamically provisioning assets when your systems are integrated into a computing grid, which will make enormously expanded assets available to users.  Processing power is turned on and off as needed, and the users – or department – pay only for the assets or cycles they use. 

Now think of what happens when the concept of “grid” expands to include “data grid,” and massively enhanced storage capabilities are also available.  Companies that take advantage of this will have a huge IT advantage over those of their competitors that do not.  And the only way such a complex system could ever be managed will be via automated allocation and de-allocation of resources.

The ideas at the heart of “on demand” seem to be right on.  After all, every manager will appreciate the opportunity not to have to pay for resources that are not being used. This as a significant step in getting the company closer to a real utility computing model.  But it just as significant that both the hardware and the software groups at IBM seem to be focused, at long last, on the same set of goals.  While the proof is always in the execution, it is worth noting that, after all this time, the various parts of IBM are once again all singing the same song, and that now it’s a song in tune with the times. 

The 1990s were a particularly tough time for IBM.  It would appear that at some point during the near-death experience of the last decade, IBM apparently saw the light.