MySQL fits Sun's strategy and ambition to deliver a top-to-bottom open-source stack. MySQL has a huge installed base, much of which uses the database in high-volume Web servers. Although MySQL was purchased for the equivalent of $1 billion ($800 million cash plus $200 million in options), its actual revenue — in the range of $75 million — indicates that Sun is paying a hefty premium. What Sun is paying for is a vast opportunity in accounts, building out large-volume Web server farms back-ended by databases. Even if the accounts are supporting their own systems, a small percentage of a large number can bring Sun a future annuity on the sales and services of database management system (DBMS)/application solutions on its hardware platforms that it otherwise would have had little hope of attaining or would have required hard sales tactics to break into the accounts. With open-source software (OSS) DBMS shipped revenue growing at a 40% rate (2006/2005), Sun has what amounts to several multipliers: hardware (x86 Intel and AMD, SPARC and Niagara), software (Linux; OpenSolaris; Windows; Apache and the Linux, Apache, MySQL, PHP/Python/Perl [LAMP] ecosystem; Java and other Sun open-source software [from GlassFish, NetBeans and so on], MySQL, Xen, Lustre for HPC) and two strong open-source communities. Sun's mantra is "value and volume."

This is not a sinister plot by Sun to hijack Linux users and convert them to Solaris. Sun's announcement overtly mentions LAMP and makes minimal mention of Solaris. Moreover, MySQL has a substantial installed base on Windows (precise figures are not available due to the freely available source code and the OSS nature of the MySQL license). It would be outright foolish of Sun to play "bait and switch" with the Linux community, whose power through community participation and action speaks louder than any single vendor's marketing message or market manipulation. Although Sun will continue to play up Solaris as a superior operating system (OS) environment to Red Hat Linux specifically (it doesn't believe Novell SUSE is competitive), Sun still has ample revenue opportunities in system integration, support and platform sales from both within its Solaris installed base (a lesser opportunity), Internet Web services (a high-volume play primarily on Linux or Windows), and new and emerging Web 2.0 businesses, which may be influenced toward Solaris if the users believe that Sun is truly an OSS vendor. However, the DBMS competitive factor is the key factor in Sun's viability and growth. Sun must begin to win the competition in future DBMS deployments related to the Internet against IBM, Oracle and Microsoft, with hopes of future replacement opportunities where these databases already exist. Bottom line: Sun will keep MySQL open as a multiplatform DBMS, although it will, nevertheless, use integration and optimization for Sun as a marketing tactic where it makes sense.

Red Hat and Novell will be under increasing competitive pressure in MySQL-based environments because Sun owns the support resources and the DBMS expertise. The Sun advantage is not immediate because there are issues surrounding integrating Sun's core business with MySQL, cross-training its engineering, development, testing and quality assurance on DBMS (for example, optimization, scaling, troubleshooting, security and architecture). Sun will have to build a Linux practice and support organization that is second to none to achieve the credibility that Red Hat has built over 15 years. There may be some internal cultural conflicts in which lifelong Solaris developers and engineers see the encroachment of Linux as career-threatening or just plain foolish because of a superiority attitude about Solaris's rich OS features. Red Hat already has Oracle to contend with at the OS support level and has MySQL asserting top-down pressure against PostgreSQL that Red Hat has aligned with IBM, HP and Dell are not likely to face any immediate danger because the vast preponderance of MySQL is non-subscription-based and could be seen as an independent marketplace. But for those moving OSS into mission-critical environments (see Note 1) and looking seriously at MySQL, Sun will be at least one of the important parties at the table. At the table, Sun will then have an opportunity to talk about its hardware and systems offerings.

We would conservatively expect that it will take Sun until at least 2010 to see any significant impact on revenue from the acquisition. Although we alluded to the volume opportunities, they were just that, opportunities. Bear in mind, MySQL can be deployed as an OSS or a commercial DBMS, and that more than 95% of the installations are not under a MySQL contract. There is nothing to suggest that this trend would have changed dramatically for MySQL without Sun acquiring it. However, with 4,000 Sun salespersons, and thousands of system support specialists, we anticipate that Sun, after a break-in training period, plus some promotions, will gradually begin to seed accounts with support contracts. Over time, a moderate percentage of these accounts could enable Sun x86 or Niagara platforms to take root. We are sure that Sun will have a financial model of the total available market opportunity, Sun-affinity opportunity and a MySQL-accelerating financial opportunity with sales quota and goals. These factors will entail many variables, such as Sun's opportunity in Solaris, Red Hat and SUSE accounts; MySQL opportunities in undeveloped areas, Oracle, Sybase, PostgreSQL, Ingres, Microsoft and IBM DB2; and Solaris/MySQL on non-Sun hardware (IBM, HP and Dell). To make definitive conclusions at this stage is premature. We will need to assess Sun's progress at the end of year one of the acquisition.

There is little question that Sun has made mistakes in the past with acquisitions. One could argue that they were made under the previous administration, but that is always a weak argument, especially when there are still many of the same executives present in the newer administration. We would argue first that all of Sun's previous acquisitions have not been unsatisfactory. Waveset Technologies for ID management is one example of an acquisition that has made excellent results. Others include Aduva (provisioning) and Afara Websystems (multithreading technology), with good results. This is the largest acquisition with Jonathan Schwartz as CEO. We believe Sun has learned much from previous acquisitions and has a strong knowledge base in DBMS.

This acquisition proves that Sun is serious about OSS and, moreover, that Sun is serious about turning the OSS vendor model into a profitable one, arguably enjoyed by few if any vendors today. Other hardware vendors, for example HP and IBM, have some offerings in OSS, but no major software products supported with subscription service. However, we believe the others will be making additional moves in this direction. Sun, however, is the first and has the largest number of major OSS offerings. In addition, Sun offers subscription support for PostgreSQL and markets a data warehouse appliance with Greenplum, a data warehouse DBMS using PostgreSQL as its underlying engine. As Sun realizes increased revenue from subscription offerings, it will prove that an OSS vendor can work. This lends credibility to Sun as a true open-source vendor. There will always be a few who believe that if any vendor receives any revenue based on OSS, it is no longer OSS; however, they have weak arguments for this.

The acquisition of MySQL by Sun adds the necessary resource to increase the maturity, stability and credibility of MySQL as a major database management system (DBMS) in the market. Initially and short term — throughout 2008 — this will have little effect on MySQL and the DBMS market.

During 2008, we believe that Sun will leave MySQL to "business as usual" while it integrates MySQL into the organization, trains its sales force to sell MySQL, trains its support staff in MySQL and assists the MySQL development team in optimization, scalability and stability of large database implementations (as Sun has done previously with DB2, Oracle and Sybase). Note that this does not only affect MySQL on Solaris but also strengthens the DBMS on other platforms such as Linux and Windows.

As Sun's sales force gains experience in selling DBMS, our view is that MySQL will gain market share against not only the other open source DBMSs (EnterpriseDB Postgres, Ingres and PostgreSQL), but also against the proprietary DBMSs (IBM's DB2 and Informix, Oracle, SQL Server and Sybase's ASE). As MySQL increases in maturity due to resources gained from Sun, it will appear more frequently in mission-critical systems (see Note 1), competing with the proprietary DBMSs. Initially, the greatest effect will be with Sybase and Oracle customers using Sun hardware, primarily due to lower maintenance costs and General Public Licenses (GPLs). An increasing number of existing customers of Oracle and Sybase will migrate to or develop new systems on MySQL.

DB2 and Informix will be less affected, as their installed base is less than that of Oracle and Sybase. Microsoft SQL Server will be even less affected as pricing is less of an issue, with SQL Server being priced well below that of DB2, Oracle and Sybase.

Ultimately, MySQL will not only gain market share against these DBMSs but will also bring pricing pressures against IBM, Oracle and Sybase. By 2010, MySQL will be considered one of the five major DBMS engines for online transaction processing (OLTP) with DB2, Oracle, SQL Server and Sybase ASE, actually pushing Sybase ASE in fourth place — based on reported revenue (see "Market Share: Database Management System Software, Worldwide, 2006").

There is no logical or rational motive for Oracle to inflame and alienate users who have existing Oracle DBMS solutions on Solaris (especially those on SPARC hardware), against Sun. Historically, there is no evidence of Oracle or any other software vendor purposely driving its customers away from a hardware vendor (although Oracle will preserve the right to pursue strategies that could be at odds with hardware vendor interests). Many of these accounts are paying substantial license and support fees to Oracle, continue to go through hardware upgrade cycles with Sun and are risk-averse to switching vendors midstream. The only opportunity for Oracle to swing users away from Sun would be for its Oracle Enterprise Linux and DBMS solutions on x86 hardware.

However, even in such cases, Sun could remain a competitive influence with x86 platforms with competition emanating from competitive hardware by IBM, HP, Dell, Fujitsu and other vendors. Many options remain open to the user, such as: Unix/RISC platform, an Oracle DBMS on RISC hardware, a Linux/x86 platform for Oracle (for example, Oracle Database 10g) or Windows running Oracle on x86. In each case, Sun has an opportunity with its own x64 systems, and with Oracle on Unix, Sun's SPARC strategy remains viable.

Remember that Sun's position is not unique, as IBM's DB2 and Informix compete against Oracle as well as HP's Neoview.

Sun will market MySQL under the same license structure and pricing that MySQL has as present. Subscription service is sold by server and not by the number of CPUs, chips or cores. With the exception of Microsoft, other vendors' pricing uses complex formulae based on cores per server.

Microsoft pricing is based on the chip, regardless of the number of cores. Additionally, with the exception of Microsoft, the other DBMS have many extra priced options, which sometimes drive the price up substantially. Maintenance and support fees are charged as a percentage of the purchase price (as high as 22%) and not as a fixed subscription fee, as with MySQL (ranging from $100 to $8,000 per year per server). Therefore, organizations are paying much higher fees for maintenance and support for IBM, Oracle and Sybase and higher (although less than the others) for Microsoft. Even where an organization has an enterprise or unlimited license for a proprietary DBMS, the total costs are considerably higher than for MySQL (even using the argument that new licenses are "free").

Vendors such as IBM, Microsoft, Oracle and Sybase will continue to see revenue based on the included functionality, which is missing in MySQL. Where this is not a factor, IBM, Microsoft, Oracle and Sybase will be forced to drop support costs substantially or MySQL will win in every situation. This will have a major impact on maintenance and support revenue of these vendors as MySQL becomes more mature and gains share in new applications. Organizations will also begin examining the cost of migration for existing applications maintenance and support costs may be low enough to provide the necessary return on investment (ROI), making it a simple decision to migrate to MySQL (see "Cost-Benefit Analysis of DBMS Migrations").

Customers of proprietary DBMSs will be able to use the lower support costs of MySQL in negotiations with the proprietary vendors, forcing greater discounts in the cost of support and maintenance across all applications. With the capabilities of MySQL to support an increasing number of applications, including clustering and data warehousing, this leverage can be used across all licenses of proprietary vendors.

We believe that Sun will lose few MySQL employees and retain the management team. MySQL has approximately 400 employees, including a strong management team. We do expect some of the developers to depart as they may now think that MySQL is not really open source, since it is now "owned" by a proprietary vendor. We disagree and believe MySQL's model entails open source but know that some of the "purists" disagree. Nevertheless, employees of MySQL, including the executives, will largely remain with Sun. This is not only a monetary decision (the acquisition included $200,000,000 in options), but also a cultural issue. The culture at Sun is similar to MySQL, providing much the same working environment that is expected. Further, MySQL CEO Marten Mickos will join Sun's executive leadership team and report to Rich Green, Sun's executive vice president of software. This will provide the necessary autonomy for MySQL to continue as previously.