Users are faced with continued media and vendor (especially IBM) hype concerning Linux in general and, specifically, its viability as a potential mainframe growth driver.

Although users have been relatively slow to embrace mainframe Linux (notwithstanding IBM's marketing figure of 11 percent of 2001 MIPS shipped being Linux-ready), we see some near-term momentum developing for small application integration (Unix application ports moving to Linux) with existing legacy systems, Linux operational shakedowns, and low-end consolidation (mostly driven by limited staff resources). Our research indicates the modest user consumption of mainframe-based Linux is largely driven by the mainframe's current ease-of-use/management lead (e.g., partitioned workloads, workload management, prioritization, ease-of-use characteristics [for VM]) versus current incarnations of Unix and Wintel-based systems, as well as the availability of some existing mainframe MIPS. But we believe few users are buying any significant new mainframe capacity to run Linux, except in limited cases where service providers are exploiting the ease of virtual machines via IBM's virtual machine facility (z/VM and VM/ESA).

[an error occurred while processing this directive]

While Linux will have only minimal effect on short-term mainframe MIPS (2-4 years), its longer-term platform effects must quickly become part of data center planning. For IBM's top customers (the 20 percent of customers that consume 80 percent of mainframe MIPS), Linux will garner less than 5 percent of all MIPS during this transition period. For the "lesser" 80 percent of mainframe customers (consuming just 20 percent of mainframe MIPS and generally well below 1,000 MIPS), Linux on z/OS will play a slightly larger "transitional" role, pointing to low double-digit adoption rates. These customers are now seeking an early start in their inevitable platform transitions (e.g., S/390 and z/OS to Unix, Win2000, and Linux) acquiring new skills on their way to significantly reduced mainframe dependencies by 2007. Moreover, by 2007, Linux will have garnered about 20 percent of enterprise operational MIPS, mostly at the expense of Unix and some Win2000. Simply put, mainframe Linux will remain a niche, whose term will be defined by the dual mainframe/Wintel gaps of cost (widening) and management functionality (closing).

Drawing the Moth to the Flame Indeed, mainframe Linux's primary attraction is as a simple "sandbox" for user experimentation, and to more closely connect existing legacy applications with outboard Win2000 or Unix solutions (which are typically more difficult to manage and, for example, schedule). More important, research indicates mainframe-based Linux's main near-term attraction is its ease of operation, not specifically its price or total cost of ownership. We believe users should view the mainframe as an interim (and potentially costly, depending how existing MIPS are charged) port in the early Linux storm. Near-term Linux mainframe deployment (through 2004) should be limited to consolidation and integration of simple applications and should be used to gain Linux expertise in a controlled, low-risk environment. Longer term (2005+), we believe the Linux center of gravity will inexorably migrate to the compelling economics and maturing mainframe-class management capabilities of Intel-based servers, leaving little justification for a continued mainframe Linux presence.

The Revenue Problem Complicating Linux for S/390 and z/OS decisions is IBM's increasingly difficult revenue goals. We do not believe IBM will be able to "afford" matching the aggressive price/performance improvement curves (35 percent/year) of the mainframe's platform competitors (especially Intel). This will put further economic distance between z/OS and Unix and Win2000 as users' Linux platform of choice (even IBM's new, relatively low-priced Linux-based z800-0LF hardware remains 8x-10x more costly than Intel-based systems, which are 4x that of Unix).

Users should leverage the significant (3x+) disparity between IBM's aggressive pricing incentives for running Linux on the mainframe against legacy workloads running on the identical hardware platform. Indeed, users acquiring low-cost ($1.6K/MIPS) Linux-dedicated mainframe capacity (e.g., the z800-0LF) should, at minimum, insist on the contractual right (and IBM's technical fixes) to use those MIPS (e.g., IPL z/OS) for legacy workloads in a disaster recovery scenario. They should further request the flexibility to shift such Linux MIPS to legacy workloads (at no additional charge) as business requirements dictate.

Business impact: Long-term Linux will afford a degree of business economic value, provided that the business and the IT organization work in concert defining realistic roles and appreciating its limitations. Linux is a small part in bringing total IT costs down.

Bottom line: Mainframe Linux's near-term attraction is it relative ease of operation, not price per se. We project mainframe Linux will remain a shrinking niche, whose term will be defined by the dual mainframe/Wintel gaps of cost (widening) and management functionality (closing).

What are your company's plans for mainframe Linux? TalkBack to us.

Newsletters
	Tech Update Today eBusiness Update	Tech Update Weekly		All newsletters FAQ Manage my newsletters

[an error occurred while processing this directive]

[an error occurred while processing this directive]