|
IBM has recently publicized its Project eLiza, a major research initiative focused on developing systems that can manage and fix themselves without human intervention. IBM expects to spend as much as 25 percent of the server group's $2 billion annual R&D budget on the effort over the three to five years. IBM has set a goal to roll out new functions every six months.
The first thing to note is that this announcement does not set a specific goal and does not set a specific deadline. Nevertheless, IBM views eLiza as addressing the industry's next grand challenge — the pressing need for self-managing and self-healing systems (or autonomic computing), caused by the dramatic growth in the number of systems and IT staffs' inability to manage those environments. Self-healing systems may not be the top priority of most enterprises today, but Gartner believes they will grow rapidly in importance through at least 2006 — particularly as skills costs escalate and exposure to downtime heightens.
In addition to self-healing, eLiza also encompasses dynamic workload management across heterogeneous systems, super-scalable clusters (potentially tens of thousands of microprocessors) and distributed server management for hundreds — if not thousands — of servers from anywhere. Although today's homogenous IBM mainframe world contains elements of many desirable properties (e.g., availability, manageability and security), the eLiza announcement aspires to embrace not just all types of servers but all types of systems that include IBM and non-IBM products. Initially, however, IBM will likely focus on a far less comprehensive goal — unifying the IBM eServer product line.
Announcing eLiza is astute public relations as it enables IBM to set itself apart in the "vision" wars from competitors such as Sun Microsystems, Hewlett-Packard and Microsoft. However, for IBM to fulfill its eLiza vision, it will need to encompass systems and software from vendors other than just IBM. IBM's ability to get agreement on standards and how control will be maintained will be particularly difficult in a market of different vendor agendas. Gartner believes that such difficulty will likely cause the eLiza vision to come to fruition later rather than sooner.
As for IBM's last big investment, Linux, to which IBM committed $1 billion in 2001, IBM views it and eLiza as complementary initiatives, although Gartner believes that the branding emphasis for eServer could shift over time.
IBM faces other internal challenges. Although IBM's server and storage groups drive the eLiza effort, other divisions — most notably, the software division — must be enthusiastically committed to the project as well. IBM Research needs to set specific goals and avoid the risk of getting stuck in the morass of trying to deliver the ultimate heterogeneous solution while smaller, more nimble competitors beat IBM to the punch with partial solutions. However, as IBM management resets some of its priorities, eLiza could serve to unify the numerous research threads that may now address only aspects of availability, scalability and manageability.
With eLiza, IBM was smart to differentiate itself from, and challenge, its competition — without making any substantive commitments. While an initiative such as this does not necessitate a change in buying patterns today, it does highlight factors that should become part of an enterprise's evaluation criteria through at least 2004. Most enterprises should hold off giving IBM any credit in request-for-proposal evaluations until the specific goals and timetables of Project eLiza crystallize. Advances will likely come in small increments rather than in one dramatic breakthrough. Type A enterprises (rapid technology adopters) that need to be at the forefront of these initiatives should enter into discussions with IBM to help shape the project's priorities.
Analytical Sources: Mike Chuba, Enterprise Systems & Centralized Operations; Thomas Bittman and George Weiss, Unix & Midrange Strategies
|
