Windows vs. Linux vs. Solaris on x86

Windows vs. Linux vs. Solaris on x86 Servers: Decision Time, Now

October 12-16, 2008 Orlando, FL

Symposium/ITxpo 2008 October 12-16, 2008 Walt Disney World Dolphin Orlando, FL

For more information about our research policies, processes and methodologies, please visit Gartner Research Methodology on gartner.com. These materials can be reproduced only with written approval from Gartner. Such approvals must be requested via e-mail: [email protected].

George Weiss


Business Value Proposition • The faster IT consolidates servers and operating

systems (OSs), the quicker the benefits in cost savings. • Most IT organizations do not have a governance policy policy on OS life cycles cycles — most will will be wasting wasting large maintenance and administration costs needlessly. • Even as an open source software (OSS) OS, users are failing to exploit Linux total cost of ownership (TCO) advantages.

IT asset management and procurement are continuing to seek ways to reduce acquisition and operational costs via open and commodity systems, but they surprisingly continue to make costly mistakes. In the early 2000 period, x86 performance and costs dramatically improved and caused a flood tide from Unix to Linux. But now, users not only have the traditional Unix and mainframe operating systems (OSs) to wrestle with, there are also numerous versions of Linux on high volumes of x86 deployments. The contract pricing on subscription services is displaying an upward creep and is looking less and less like the bargain that migrating workloads to x86 was meant to be. Without governance, policies, standardization, consolidation and OS life cycle modeling, IT organizations may be diluting and negating much of the cost savings they should have inherited. This presentation looks at the OS options, how they will compete and what to do about making the selection process turn into a positive experience.

This presentation, including including any supporting materials, is owned by Gartner, Inc. and/or its affiliates and is for the sole use of the intended Gartner audience or other authorized recipients. This presentation may contain information that is confidential, proprietary or otherwise legally protected, and it may not be further copied, distributed or publicly displayed without the express written permission of of Gartner, Inc. or its affiliates. © 2008 Gartner, Gartner, Inc. and/or its affiliates.Al l rights reserved. reserved.

George Weiss SYM18_1062, 10/08, AE

Page 1


Key Issues • How will the market dynamics among OSs affect

the selection process? • What are the vendor and OS strengths,

challenges and critical success factors for x86 servers? • How should users optimize, negotiate and

manage the OS life cycle?

This presentation, including any supporting materials, is owned by Gartner, Inc. and/or its affiliates and is for the sole use of the intended Gartner audience or other authorized recipients. This presentation may contain information that is confidential, proprietary or otherwise legally protected, and it may not be further copied, distributed or publicly displayed without the express written permission of Gartner, Inc. or its affiliates. © 2008 Gartner, Inc. and/or its affiliates.Al l rights reserved.


Page 2


Key Issues • How will the market dynamics among OSs affect the

selection process? • What are the vendor and OS strengths, challenges and

critical success factors for x86 servers? • How should users optimize, negotiate and manage the

OS life cycle? Windows and Linux will be the leading OSs on x86 by a wide margin — Solaris will be an important but distant third during the next 24 months. As the race intensifies, the costs of supporting Linux, Windows and Solaris on x86 will be negotiable and more competitively aligned for similar systems.



Page 3


x86 Servers Will Propel Windows and Linux as the Two Primary Growth Operating Systems Vendor Revenue ($)

x86 Server Shipped Server Revenue by OS

Linux Unix Windows

30,000,000,000

Other

25,000,000,000

Windows (9.3% growth)

20,000,000,000 15,000,000,000 10,000,000,000

Linux* (12.7% growth)

5,000,000,000 Unix (0.8% growth) Other (6.3% decline)

0 2007

2008

2009

2010

2011

2012

2013

* Note: Server revenue associated with Linux is higher in reality due to a high proportion of unsubscribed Linux. Unsubscribed rates may be as high as 40% of total installations.

The Unix segment will have a declining market share with a compound annual growth rate (CAGR) of 0.8% in the next five years. HP-UX, IBM AIX and other proprietary Unix systems will be shrinking at CAGRs of 1.6%, 0.4% and 6.3%, respectively. The shrinking market size of these segments is due to the continuous substitution with less expensive alternatives, such as Windows and Linux, and more expensive support resources. Sun's growth strategy for Solaris on x86 and its marketing of Solaris as a better OS (compared to Linux) on multicore servers has helped give the category a higher-than-average growth rate within the Unix segment. The Windows server segment will grow at a 9.3 percent CAGR through 2013. The adoption of the new Windows 2008 Server OS could be as slow as that of Vista. Enterprises on average take at least 12 months to prepare for the migration, and they consistently wait for the first Service Pack to be delivered before adopting a new version of Windows. In some cases, users delay the migration until earlier versions of Windows Server are no longer supported by Microsoft. The Linux server segment is the fastest growing server OS market, having a CAGR of 12.7 percent. The cost-toperformance ratio, platform openness and increasing availability of applications are key drivers for the aggressive adoption of Linux. More and more vendors are offering Linux as an alternative to legacy or proprietary OSs (in particular as Unix replacements) in selling hardware and suite solutions. There is also a trend that more Linux is being used in mission-critical systems.



Page 4


Windows and Linux Are Dominant on x86 and Niche on Non-X86; Solaris Is Positioned to Exploit Both x86 OS Ratios Modular blades and racks from inexpensive components

Revenues ($B) Windows 20.4 Linux 7.4 Unix 1.2 Other 0.8 Total 29.7 2007

2007

Revenues ($B) Windows 0.36 Linux 0.92 Unix 15.7 Other 8.1 Total 25.1

Linux Windows

25%

68%

non-x86 OS Ratios

Other

Inversely Positioned OS Opportunities

32% Unix 63%

Vendor-customized frames for low latency, high availability

Server architecture influences OS selection. The technology can be bifurcated into two main streams — x86 and non-x86 — including Itanium and reduced instruction set computer (RISC) variants. On a revenue basis, they are within a few billion dollars, but on a unit volume basis, x86 completely dominates. This makes the selection of the OS a vital and long-term decision. Three primary OSs (Linux, Windows and Solaris) can run on x86 hardware (Linux and Windows on Itanium, but niche). But Linux consists of a variety of alternatives from Red Hat, Novell, Oracle, Ubuntu, CentOS and many others, plus numerous back versions on legacies. Windows is in the midst of its own upgrade and has been facing stiff challenges from Linux — especially in Web servers that could extend into the cloud. Solaris is open source and is offered freely downloadable as a subscription service, but it is late to capitalize on the open source momentum. Within the non-x86 market, there are three leading Unix OSs, the IBM z/OS system and others — with most of these vendors competing seriously for IT deployment share. This has created competitive dynamics in which most Unix server vendors are actually competing against and among themselves since they can deliver Windows, Linux and Unix. But independent software vendors (ISVs) — for example, SAP and Oracle — are now turning against Unix, realizing that it’s more important to help users reduce their hardware costs and prop up their own software licensing. Inevitably, the migration that started in 2000 will continue and will favor Linux and Windows while Solaris, as the only x86 Unix, attempts to add share. This presentation, including any supporting materials, is owned by Gartner, Inc. and/or its affiliates and is for the sole use of the intended Gartner audience or other authorized recipients. This presentation may contain information that is confidential, proprietary or otherwise legally protected, and it may not be further copied, distributed or publicly displayed without the express written permission of Gartner, Inc. or its affiliates. © 2008 Gartner, Inc. and/or its affiliates.Al l rights reserved.


Page 5


Windows and Linux Will Continue to Capture Most New Workloads — Showdown to Come in Clouds June 2008 poll: Compared to 2007 levels, do you anticipate that the total workload for your organization’s server environment will increase, stay the same or decrease in 2008?

Windows Linux HP-UX Solaris AIX z/OS

Increase 81% 71% 26% 39% 48% 33%

Decrease 3% 2% 44% 29% 18% 27%

Same 16% 27% 39% 32% 34% 40%

Number of Responses 115 67 27 34 44 30 n = 115

While the poll in June is not a statistically scientific finding, we can still draw some conclusions because of the stark contrasts. New workloads are dramatically more likely to be aimed at x86 (and, therefore, Windows and Linux) than any other OS environment. The only Unix system that compares favorably is IBM’s AIX with Sun, HP and IBM’s z/OS systems trailing. From these polls, Gartner has concluded that Linux and Windows will be the two primary growth OSs, driven in large part around lower acquisition costs and maintenance on high-end systems. Of course, other systems will not disappear due to plenty of new and upgrade hardware Unix contracts amounting into the million or multimillion dollar range. The battlefield will then be won on the basis of performance, efficiency, utilization, agility, availability and security. But as virtualization brings many of the same features of Unix to x86, such as mobility, high availability (HA) and fault tolerance, capacity planning and placement optimization, it becomes increasingly evident that the OS is becoming more divorced from the underlying bare metal and more distributed in nature. This will make the OS decision more dependent on external infrastructure variables than just on the features and functions tied to the kernel.



Page 6


Nonzero Sum Market Dynamics Assures That Linux Momentum Continues for the Next Five Years Linux has a major opportunity in OS decisions: IT data centers will face OS consolidation, often at an intersection with Linux.

Compete + expand Linux market Compete + expand SUSE opportunity Compete (Unix) + expand Linux market Expand MFs and System p + Linux opportunity Compete + expand market Compete (Solaris) + expand (OSS)

Nonzero Market Dynamics

In a nonzero sum market, both competition and collaboration are necessary ingredients to survival. A greater number of opportunities to partner may broaden market opportunities that otherwise might not have existed. For Linux in particular, since it is not a single, one-of-a-kind packaged product — and it can be an enabling technology in a multiplicity of markets and vendor products — the opportunities to put down roots are manifold. The table lists combinations in which vendors may appear to be competing but also are contributing to the further growth of Linux. Whether it’s Red Hat vs. Novell, or Novell vs. Microsoft, or Oracle vs. Red Hat, or even IBM vs. IBM, or Sun vis-à-vis open source, in each case there is a nonzero sum in which Linux benefits. This is why we can declare that the tipping point has already been achieved toward continued growth and market gains. Later, we will discuss OS life cycle with respect to OS maturity and ISV commitments as additional factors in measuring the success of an OS.



Page 7


Windows Server 2008 Delivers Important but Incremental Improvements Over 2003 Pros • Windows Server 2008 delivers

important improvements over 2003 • Enables virtualization (Hyper-V) • Enables drive to x86 64-bit

computing • Server core for specific case uses:

physical and virtual appliances, Web farms, blades

Cons • Improvements are incremental

(vs. innovative) • Late to 64-bit computing* • Late to virtualization* • No real-time kernel* • Late to HPC clustering*

• Stepping stone to a more modular

kernel

*(compared to Unix or Linux)

As with all OS vendors, Microsoft continues to enhance its products in areas of improved performance, better security, increased manageability, virtualization and best-in-class reliability and availability. We believe that for Microsoft to keep Windows competitive, it must show that it can rapidly innovate and assume leadership in as many parts of the market as possible — from high-end OLTP database management system (DBMS), business intelligence, analytic processing, scientific and engineering algorithms, real-time applications in financial services and telecommunications and so forth. We recognize that Windows Server 2008 has literally hundreds (if not thousands) of granular improvements over Windows Server 2003 and Windows Server 2003 R2. But many of these improvements reside at a low level in the kernel. Thus, we see Windows Server 2008 as an incremental improvement over Windows Server 2003, not a major shift. An incremental improvement should not be regarded as a negative; most IT organizations prefer incremental improvements that can be easily digested instead of large technology jumps that require massive changes in existing infrastructure. But we also see a lot of room for improvement, such as in clustering/HA, real time, virtualization and 64-bit applications. These are environments that would extend its envelope of influence beyond the back office where SharePoint, SQL Server and IIS have been the main driving forces of Windows Server.



Page 8




critical success factors for x86 servers? • How should users optimize, negotiate and manage the OS

life cycle? 158875 Toughness ID:G00 d an y t ili Ag t ke ar allenge: M Red Ha t's Ne x t Ch How Micr osof t, Su n and Nov ell Can Change Linux Mar ke t Dy namics ID:G 00152637 Key Issues for Servers, 2008 ID:G00156426 606 :G00156 ID s m e t s os y : S un M icr g in t a R r o V end Nov ell' s Challe nge: Ex ecut ing on It s P ot ent ial ID:G00147 6 58



Page 9


All Operating Systems Will Face Challenges to Gain Leadership in the x86 Server Market

Key Issue: • Can Microsoft broaden appeal to massively scalable infrastructures (grids, HPC, clouds)?

Key Issue: • Can the Linux distributors do for virtualization what they did for physical?

Challenges: • OSS availability and licensing appeal to nextgeneration data centers • Negligible migrations from Linux to Windows • Driving success in virtualization against VMware

Challenges: • Lack of clarity in creating a virtualization management ecosystem • Above the kernel values may create new lock-ins • Subscriptions getting pricier

Key Issue: CanIssue: Sun disrupt both the •Key and Linux • Windows Is server virtualization markets core? without • disrupting $500M foritself? what? • VDI is focus Challenges: •Challenge: Monetizing OSS into revenue Lack of clarity in their • predictable streams strategy (hear one thing from Citrix, into another the Linux from • Wedging market XenSource) Lack of commitment Solaris •• Re-inspiring developers competing with and Microsoft users • Basically, unsure futures.

Users will face as much OS decision difficulties in the x86 market as in the non-x86 server market although the cost, flexibility and modularity in hardware mitigates some of the issues. The competitive issues revolve around business models, feature/function velocity, physical to virtual transition, platform and OS management and coping with the extraordinary high volume potential of virtual machines (VMs) and clouds. How and when to use open source, what subscription licensing is appropriate, how to create or re-create mission-critical infrastructure and business continuity all matter. It all comes down to how much can an IT organization leverage and harness its current infrastructure as it transitions into new architectures, new licensing terms, multivendor integration and interoperability, vendor accountability, service levels and indemnification. Overconfidence can lead to higher risk while too much caution can slow an enterprise from making rapid strides toward IT modernization. In developing an IT modernization strategy, IT planners should become familiar with the main key issues and challenges facing the x86 OS vendors detailed in the chart and should checkmark compliance and strategies on a risk measurement scale. For example, if Solaris is wedging into the Linux and OSS market, by how much and how fast? If Red Hat intends to lead in virtualization, where are the proof points and architectural underpinnings?



Page 10


Windows Market Outlook: Exploit, Expand and Evolve Strategies for market expansion and growth: • Exploit new market opportunities in clustering, HPC, cloud and virtualized

infrastructures • Expand portfolio with OSS applications - Test, certify and optimize OSS for Windows

• Expand developer loyalties (.NET) and Linux interoperability - Test and certify Novell SUSE Enterprise Linux interoperability - Use improved interoperability to discourage migrations from Windows

• Reinforce IP through indemnity agreements • Evolve virtualization with strategic emphasis on comprehensive physical and

virtual management Most optimistic: Slows Linux growth through hybrid Windows/OSS

Scenarios Least optimistic: Fails to slow OSS/Linux in cloud, licensing model and OS modularity

Most likely: Retains midupper tier back-office strength, but lacks force in clustered, Web-centric infrastructures

The major challenges in Microsoft’s future will be its credibility and accomplishments in delivering cloud computing infrastructures and services along with virtualization. It needs to address the market’s perception of late arrival and limitations in clustered and massively scaled servers. As one important example, Microsoft lacks the technology and market initiative that Oracle has assembled with its Oracle DBMS Real Application Clusters (RAC) for Linux on x86. Another factor in the market affecting Microsoft is open source software (OSS). If open source is a tactic to gain share among Linux distributors, then Microsoft can use these packages on Windows to expand the relevance of Windows into markets that would have been closed. Its strategy with Novell supports interoperability, especially in a virtualized world, where it must deliver integration and not isolated or self-contained Windows solutions. This is why the investment in Novell has continued unabated despite some cynics that believed Microsoft would soon abandon the relationship. While it seeks integration points at the server level, it has other ambitions, such as hosting virtualized desktops, that will be a major hurdle for Linux distributors like Red Hat who have no working Microsoft relationship. Thus, a one-two-three punch by Microsoft would be: (1) OSS on Windows, (2) strongly managed server virtualization and (3) application hosting and streaming from Hyper-V virtualized servers.



Page 11


Sun Solaris Market Outlook: Driving an Open Source Model Into High Returns Strategies for market expansion and growth: • Drive innovative and energy-efficient product design and packaging • Re-energize the developer community - Use OpenSolaris as a better Linux alternative while exploiting advantages of OSS - Seed loyalties bottom-up through OSS development - Shift company to higher growth through OSS appeal in developing economies

• Shift focus and design to high-volume and massively parallel markets and

infrastructures using the subscription support model of L inux Scenarios Most optimistic: Slows Linux; revitalizes Solaris; generates large pipeline of systems and services

Least optimistic: Fails to staunch the Linux appetite and ongoing Unix migration

Most likely: Takes two years to achieve a stable, critical mass in volumebased subscriptions

Solaris is the other major competitive alternative to Linux. Sun has been OSS friendly and continues to drive OSS up the stack (for example, MySQL). But its initial positioning for Solaris on x86 was designed as a zero sum takeaway from Linux. Ultimately, the message played poorly among the Linux aficionados while it was finding challenges in winning back market share in its traditional markets. Today, Sun has modified its strategy and messaging that highlights the benefits of a better Linux but leaves plenty of wiggle room for users to find an optimum combination of Solaris and Linux across multitiered architectures. Sun’s biggest hope is to continue seeding a large number of the next-generation IT data centers in need of strong system and software support so that Linux distributors will be resource challenged. But even as it tries to build OS differentiation, many users are more interested in a modular OS to minimize the cost of operations and perhaps enable some customization. Sun hopes the market will recognize that like Linux, OpenSolaris will provide identical provisioning, patching and subscription support but with superior scaling, virtualization and system building blocks for massively scaled infrastructures. Its biggest challenge is the rapid conversion of new Solaris development projects to commercially supported accounts in a business model that can keep IBM and HP at bay while building a large new business in emergent IT and growing world economies.



Page 12


Linux Market Outlook: More Than "Linux" — Red Hat, Novell, Oracle, CentOS, Ubuntu… Strategies for market expansion and growth: • Appear unified while competing for share growth and migrations • Drive next stage expansion beyond Web and application servers to critical

applications

- Oracle RAC, WebSphere, JBoss, real time, HA, mainframe, massive scale - Be perceived as low risk, high value

• Simplify upgrade, maintenance and support while adding functionality,

expanded hardware compatibility and system management • Drive toward a strategic data center position on virtualization Scenarios Most optimistic: Moves briskly to higher-tier Unix displacement and Windows migration

Least optimistic: Noncommunity-spirited battles that re-create old Unix wars of proprietary flavors

Most likely: Eliminates barriers to entry; bifurcates into unpaid, internal or outsourced support or value-based pricey support

Technically, it is not correct to analyze the third pole of the x86 market as Linux. There are about 100 Linux distributions, any of which a user or IT organization can download for development or production use. And the kernels and tools are also widely divergent — enough that an ISV cannot simply certify an application on one OS and expect it to run error-free on any other. One of the major ISVs clearly stated to us that wide variations exist from one distribution to the next, from one version to another. Red Hat and Novell make decisions to integrate features and capabilities that may have been only lightly tested by the Linux kernel community into their individual versions at different rates and at varying quality levels. This factor is probably one of the key variables in why so many IT shops steadfastly cling to Unix, Windows and mainframes. It takes only one bad experience in a mission-critical instance to sour an IT organization about the risk of pervasively running their systems on Linux. There’s always the temptation for a vendor or ISV to get into the act and shape up the Linux kernel for specific roles. While any vendor can do so, the fear by users that vendors will begin another forking contest like Unix and create market fragmentation inhibits vendors. However, Oracle has shown that a vendor can step up to Linux support, show that Unix is no longer mandatory in the DBMS tier and deliver solid tierthree level support.



Page 13


DBMS Platforms on Unix Decline as Linux Rapidly Increases Worldwide Linux RDBMS Software Vendors for 2007 Based on Total Revenue (Dollars in Millions) Vendor

2006

Oracle

1.954.5

2.933.9

2.294.3

3.712.5

Total (all vendors)

2007

Market Share Market Share Growth (%) (%) 2007 (%) 2006 2006-2007 85.2

79.0

50.1 61.8

Worldwide Unix RDBMS Software Vendors for 2007 Based on Total Revenue (Dollars in Millions) Vendor Total (all vendors)

2006 5.302.3

2007

Market Share Market Share Growth (%) (%) 2007 (%) 2006 2006-2007

4.972.8

-6.2

Two separate issues are evolving from the open source world: open source DBMS engines and Linux. In 2005 to 2006, Linux picked up speed while Unix declined in total 2006 revenue. This trend is accelerating and will continue. In 2008, Linux’s DBMS revenue will be approaching 30% to 40% of total DBMS license revenue. As Linux becomes "enterprise ready", with a proven track record, robust management tools and robust best practices, vendors will gradually diminish support of the Unix OS. IT organizations should have been developing Linux expertise and deploying Linux as an application platform wherever possible. All of the major relational DBMS vendors are embracing Linux and even switching their primary development platforms to Linux (Oracle began with Oracle 9i). Today, there are still some issues with the Linux environment and its ability to support a true production environment, but these issues are being addressed by all the major vendors and will diminish in the next two years.



Page 14


All 3 OSs Have Strengths and Vulnerabilities: Tipping Point May Hinge on 1 or 2 Factors SWOT Strengths

Linux  

Major OEM support Flexible HW/SW options

  



Weaknesses

Opportunities

 

  



Threats

 

Data center proof points Migration, services

Cloud infrastructures HPC and real time Virtualization

OpenSolaris Windows + OSS Lack of portability

Solaris

Windows



  

Partner ecosystem Integrated stacks Developers New feature time-tomarket Heterogeneous interoperability

Cloud infrastructures Virtualization HPC

 

 

 



  

Richness of Unix Developers

Financial consistency Challenging Linux

Emerging economies OSS projects (for example, MySQL) Virtualization

MySQL, Oracle RAC  Data center migrations Unix to Linux migrations to Linux Mainframe Linux  Red Hat/Novell/Oracle consolidation * exception may be Sun Solaris for x86

In this "super-SWOT", we have sought to characterize the strengths, weaknesses, opportunities and threats on an aggregated basis across the three major environments. While it is possible for all to partake in the emerging new opportunities because of the compatibility offered on x86 platforms (chip-level feature/function enhancements), each of the vendors and environments bring a set of unique differentiators. Moreover, an OS is just one element of an increasingly complex stack of software elements. Interchangeability among vendor platforms is increasingly more difficult — even though in one case, Sun Solaris, is available on Sun x86 hardware as well as IBM, HP and Dell. Applications used to be the major driving force in an OS decision. This will change with the increased interoperability of VMs across x86 and the abstraction of the OS and applications from the hardware. The resultant impact is that the value proposition has moved up the stack, and the decision point has moved with it. The kernel richness diminishes in importance to the meta-OS features that allow the OS to function as a part of a distributed resource environment. As IT resource planning becomes more complex in heterogeneous data centers, users will depend on the broader ecosystem to drive platform decisions. The return on investment (ROI) gain will be crucial in the users’ willingness to migrate off current platforms or expand their current platform usage. That decision’s tipping point may hinge on just one or a few significant deliverables among the ones shown in the chart.



Page 15




critical success factors for x86 servers? • How should users optimize, negotiate and manage the G

OS life cycle?

N I N N A L P

Growing

N

Emerging

O I T A

g i n g r e E m

g g i n t i n a w i d l r o o G s o n C


r e t u a M

g i n i n l e c D

R G I M


Page 16


Use the ISV Enthusiasm Model to Gauge OS Acceptance, Momentum and Life Cycles

o i l o f t r o P e r a w t f o S f o y t i l a u Q d n a e z i S

HP-UX (PA-RISC)

Windows (x86)

•

Solaris (SPARC)

•

Safe for many target segments and installed base leverage

IBM AIX

•

HP-UX (IPF)

Red Hat (x86)

•

Novell SUSE (x86)

• Solaris (x86)Other Linux Linux (Power)

Safe for virtually all mainstream deployment

(x86)

•

Independent Software Vendor Enthusiasm

Limited verticals, niche segments and installed base only

Portfolio and use skewed toward application, Web and infrastructure serving Portfolio and use skewed toward DBMS and verticalized workloads Broad variety of workloads

ISV enthusiasm is currently a major driver behind OS success. This enthusiasm helps generate demand for widely adopted products, which in turn, stimulates yet more ISVs to increase their focus. So a natural cycle is created that both maintains the leadership edge of successful OSs and can also create a barrier to entry for others. Some OSs will only address a niche audience but can still achieve some market success with a limited set of applications. Added complexity is created by binary compatibility issues between like OSs. Linux, for instance, runs on many platforms, but software compatibility is limited according to architecture. We have charted the relative size of a software portfolio on the vertical axis against the enthusiasm and popularity among the ISV community on the horizontal axis. In addition, the shapes signify whether an OS has a diverse and balanced portfolio that suits multiple styles of application or those cases where the portfolio is geared mostly toward network-centric workloads (application, Web and infrastructure serving) or data-centric workloads (DBMS and vertically scaling applications). It’s instructive to note that much of the viable green territory of the chart is the habitat of OSs on x86, or at least the intended target (note the arrow pointing up and to the right for SUSE Enterprise Linux and Solaris on x86). Note also that the Unix positions are still in safe territory but are and will be subject to downward forces — that is, diminished interest by ISVs.



Page 17


Use the OS Life Cycle Model as a Harbinger of Risk < SWEET SPOT >

Windows Red Hat Linux SUSE Linux

COLOR KEY

G

2007

N I

2009 2011

N

x86

N A L

Solaris HP-UX AIX

P

Growing

N O I T

Emerging

RISC/Itanium

z/OS

A

i5/OS

R

OpenVMS

G I

Netware Relative New Deployment Risk

M

g i n g e r m E

g i n w r o G

g i n t a i d l o s o n C

r e t u a M

g i n n i l e c D

OSs and their associated platforms can be classified according to five categories based on their maturity and market acceptance. The five main categories include: Emerging, Growing, Consolidating, Mature and Declining. Each category is further defined by a set of management processes that enable the IT architecture committee to qualify each OS by usage and life cycle. Using this model as a Decision Framework, enterprise architects and planners can minimize risk, such as platform/OS obsolescence, lack of support or higher support costs. We recommend that architects and planners do an inventory analysis of their production OSs and consolidate them to a few strategic OS/images for mainstream enterprise applications and solutions. We recommend that IT asset managers analyze and review annually each OS to determine whether it remains in the current category or should be moved from one category to another. Prepare to phase out within 12-24 months an OS determined to be in the Declining category and phase in an OS that is considered in the Growing category.



Page 18


As Commitments to Linux Run to High Volumes, Negotiate Aggressively and Consider Self-Support Case Examples User A pays $200/server/year on high-volume low-end tier distribution but $1000/server/year on high-end (qty 4000) — Negotiate an enterprise agreement User B was priced by the Linux distributor at $1200/server/year and by the platform vendor on same Linux distribution at $700/server/year (qty 300) — Consider IBM, HP User C found that reorganizing a large DBMS on a rack of 20 or more x86 servers was more expensive than the current single midrange Unix server license (qty 1) — Remain on Unix User D Linux distributor charged by server for a few images — Use a clone; enforce standardization and governance; hire Linux experts and manage support with help of the Web

We have seen an increase in the number of inquiries concerning Linux subscription license fees as deployments escalate into the hundreds and approach the thousands of servers. While consolidation is occurring around Windows and Linux on x86, there are still a number of options with regard to negotiating favorable contracts. For example, some users are opting out of support contracts on lower-tier deployments and where they have improved overall Linux administration skills internally. Others have outsourced support contracts, and yet others have vested support with the hardware platform provider, such as IBM, HP and Dell. The more that IT organizations can consolidate server deployments and OSs to fewer standard and cloned images, the less likely the need for a large number of disparate support contracts. The hardware vendors who buy original equipment manufacturer (OEM) subscription support contracts can offer discounts as high as 50% off the distributor’s list. Competitive takeouts, such as Oracle replacing Red Hat or Novell in place of Red Hat, can also offer significant discounts. In some cases, a Windows versus Linux bake off can offer substantial savings as long as the IT organization has flexible options on either OS. Finally, Linux may not always offer the expected savings over Unix when migration charges are factored in, such as complex databases that are already tuned and perform well on single symmetric multiprocessing (SMP) machines. The DBMS on clustered x86 nodes may require more central processing unit (CPU) licenses at lower scalability than an SMP, making the Unix midrange a better TCO option. This presentation, including any supporting materials, is owned by Gartner, Inc. and/or its affiliates and is for the sole use of the intended Gartner audience or other authorized recipients. This presentation may contain information that is confidential, proprietary or otherwise legally protected, and it may not be further copied, distributed or publicly displayed without the express written permission of Gartner, Inc. or its affiliates. © 2008 Gartner, Inc. and/or its affiliates.Al l rights reserved.


Page 19


When and What Migrations Make the Most Sense? From

To

Windows

Solaris

Solaris • Employ Sun OSS stack (for example, MySQL)

• High license fees • Large replicated server farms • Compute clusters • OSS tools/apps.

Windows

Linux

Linux

• Inexperienced admins. • Bad experience/ support • Proven back-office stack • New applications not available on Sun x86 servers • Consolidating away from Unix

• Poor Linux experience • Higher performance • Mission-critical support

• SPARC end of life • Standardizing on non-Sun x86 • Migrating Unix to Linux • Building Linux skills

With x86 as a popular common denominator technology for infrastructure, Web, application and even DBMS servers, many IT organizations face a transition from legacy platforms running on non-x86 technologies or older x86 platforms. They are concerned about migration costs but have no preconceived plan of which applications should be migrated, when and in what time frame. The matrix is meant to be a simplified approach to beginning a planning process toward making migrations by examining the decision variables, pro or con, in specific migration source-target combinations. As an example, a Windows-to-Linux source-target option would be considered acceptable when Windows license fees are inordinately high, large replicated server farms or compute clusters could enable free self-supported downloads and preferences exist for open source tools and applications. In contrast, a Linux-to-Windows migration might be preferable when a bad experience in support or implementation was painful or the IT organization is insufficiently staffed with Linux skills but tried Linux anyway. Similarly, Solaris and Linux migrations can be validated against the missioncritical nature of the applications, the organization’s willingness to standardize on Linux and diminish Solaris (or vice versa) and to ramp up Solaris as a better alternative to Linux on Sun hardware. Planners are urged to create a migration planning document for all relevant combinations as guidelines for x86 OS deployments.



Page 20


Recommendations and Actions 

Today: Use the life cycle, ISV, migration and SWOT models to validate OS strategic planning.



Use x86 as the common denominator to enable more aggressive price and support negotiations.



Tomorrow: Begin an analysis of platform and OS use cases and bracket each into best-to-least fit for purpose.



Long Term: Get familiar with the vendors’ plans to innovate, integrate, support and manage their OSs into massively scaled infrastructures for virtualization and cloud-based computing.



Page 21

Windows vs. Linux vs. Solaris on x86

Recommend Documents