Well, we find ourselves at the midway point of another year as we are now officially half way through 2009. And like we do each quarter, it is time for us to take a look at the activities within the
database market over the past three months. This past quarter was very active, with acquisitions (including one huge acquisition surrounded by rumors and industry giants), announcements, and more
layoffs due to the economic downturn, among other “stuff.” So let’s dive right in and see what happened during April, May and June of 2009.
The Sun Sets on IBM as it Rises on Oracle
There is no way we can start this edition of The Database Report without beginning our coverage with the colossal acquisition of Sun Microsystems that went
down this quarter. Over time it became evident that Sun Microsystems – server manufacturer, owner of MySQL and originator of Java – was unlikely to be able to survive and thrive as an
independent company. Its server sales were lagging behind HP and IBM, and Sun as a company seemed to be confused as to its future direction. Over the past couple of years, industry pundits ruminated
over which companies might be able and, more importantly willing to acquire Sun. This talk heated up considerably early in the second quarter of 2009, with the early favorite being IBM to take over
ailing Sun.
In late March 2009, the Wall Street Journal reported that IBM is in talks to acquire one of its arch-rivals in the server space, Sun
Microsystems. Now there were a lot of reasons why IBM would be interested in Sun. The server business is obviously Sun’s biggest business, and it would benefit IBM to expand in this area by
taking out a rival (Sun), growing against a large rival (HP), and positioning against a new entrant (Cisco, more on them in a moment). Additionally, one of Oracle’s largest hardware platforms for
their DBMS is Sun, so if IBM acquired Sun it could give them some leverage in dealing with Oracle. Although IBM has a version of DB2 that runs on Sun’s Solaris operating system, it has not been
widely adopted or implemented. IBM’s Informix DBMS actually has a bigger footprint on Sun than does DB2.
Sun has also been busy offering cloud computing and storage offerings, which is something that IBM has not yet aggressively pursued. Acquiring Sun would offer IBM a shortcut into that business.
But if you look past the server side of the equation, Sun offered some interesting technology, hardware and software that IBM was probably very interested in. A couple years ago Sun bought
StorageTek, and IBM’s very healthy storage business would be happy to add Sun’s storage business. Last year Sun bought MySQL, and IBM’s data management group would be very happy to add the number
one open source DBMS to its line of offerings. IBM tried to enter the open source DBMS fray a couple years ago with its Derby/Cloudscape offering, but it has not been very successful. And let’s not
forget than Sun created Java, and IBM would be happy to add that feather to its software cap.
The bottom line would be the price of the deal: how cheaply can IBM snag Sun? And would there be any other suitors? And what about Cisco entering the server business? Ill-advised, in my opinion.
Cisco partners with all of the server providers to sell its networking hardware. Does Cisco really want to take the risk that HP and IBM would get irked at them for competing in the server business
(no matter what the public stance of the companies) and have them more aggressively partner with a Cisco competitor? And even if you look at it from just an internal Cisco perspective, the server
business is a mature market with very large incumbent players. How many Cisco customers would rather have a server from a brand new player (Cisco) rather than an entrenched, solid company like IBM or
HP? Some, perhaps, but I doubt enough to make Cisco a credible player in this market.
But back to IBM and Sun for a moment: If this acquisition could have been closed, there would have been immediate, loud opposition… most likely with Microsoft and HP leading the chorus saying this
would be too much of an IBM monopoly in servers and cloud computing. Oracle, too, might join in worried that IBM would boost the capabilities of MySQL so much that it would drive down the price (and
earnings) of Oracle’s DBMS.
And antitrust experts predicted that IBM would have a difficult time getting the deal cleared by both the U.S. Department of Justice and the European Union because of the many overlapping product
lines. Problem areas cited by these experts included Sun’s tape-based storage systems (giving IBM a near monopoly in that market) and Unix-based servers (giving IBM about 65% of that $17 billion
market according to figures from market research firm IDC).
It would have been interesting to watch. But it never happened. The reported price tag for IBM to buy Sun was $6.5 billion. That sounds like a nice price for IBM given all they stood to gain. But not
all parties could agree on that price tag.
In early April, the talks between IBM and Sun broke down as Sun’s board rejected a formal offer from IBM to acquire Sun, which then led IBM to withdraw its offer. But rumors still swirled that
something might yet happen between IBM and Sun. When the news broke that the deal was off, Sun shares sunk by $1.90 the next day, off 22% to $6.59 a share.
Evidently Sun’s board was split over whether to approve the deal or not. News leaked that Sun’s chairman and co-founder, Scott McNealy, opposed the deal, whereas Chief Executive Officer Jonathan
Schwartz favored the deal. And it was also leaked that price was not the overriding factor that killed the deal. It seems that the group siding with McNealy felt that IBM had too much leeway to walk
away from the deal.
Around the same time as the IBM deal died, news started circulating that Oracle and Hewlett-Packard might have made a joint offer to acquire Sun Microsystems. Evidently Oracle would have snagged
Sun’s software portfolio for $2 billion, and HP would acquire the hardware business. So even as one deal died, the rumor mill kept swirling… and, as it turned out, rightly so!
But then things got quiet for a couple of days. And just when it seemed like nobody was interested in acquiring Sun after the IBM deal fell through, Oracle started singing “Here Comes the
Sun” and swooped in with a bigger offer on April 20th. The price for Sun? $9.50 per share, or approximately $7.4 billion.
Although an IBM acquisition of Sun probably made more sense, there are definite positives for the Oracle deal. Both Oracle and Sun sell mostly to larger enterprises. Many of Sun’s servers run
Oracle software – both the DBMS and the applications. By adding Sun hardware to the mix, Oracle can now offer completely integrated systems – including server, storage, OS, DBMS,
middleware and applications – without having to involve any other vendor. In point of fact, Larry Ellison highlighted this very notion saying, “Oracle will be the only company that can
engineer an integrated system – from applications to disk – where all the pieces fit and work together so customers do not have to do it themselves.” Not even IBM can claim that.
What else does Oracle get by acquiring Sun? Well, it gets a huge piece of the open source DBMS market by adding MySQL to its DBMS portfolio. Remember way back in the fourth quarter of 2005? That is
when Oracle acquired Innobase. Innobase was the maker of InnoDB, an add-on storage engine for MySQL. InnoDB is distributed under the GNU GPL open-source license. InnoDB is one of several storage
engines used by MySQL for handling more complex transaction processing requirements. Now Oracle “controls” both Innobase and MySQL.
Oracle also later acquired Sleepycat Software another provider of open source DBMS software. Berkeley DB is the open source, embedded database software offering of Sleepycat Software. Berkeley DB is
pervasive in the open source world as it is embedded in several popular open source products including Apache web server and the OpenOffice productivity suite, among others. Oh, and by the way, with
the acquisition of Sun, OpenOffice is now under Oracle’s purview. So we see Oracle becoming not just the biggest DBMS software provider on the planet (open source or otherwise), but also
expanding its open source coverage.
Still on the DBMS area for a moment, it will be interesting to see what steps Oracle takes to bolster its “database machine” strategy. Late last year Oracle announced the Oracle Database
Machine, touted as the world’s fastest database machine. It boasts eight Oracle Database servers on 64 Intel processor cores running Oracle Enterprise Linux and Oracle Real Application Clusters. In
the first quarter, Oracle sold its first Oracle Database Machine. Will Oracle adapt or augment that strategy with its newly acquired Sun Solaris hardware? I don’t know why they
wouldn’t.
Cloud computing is another interesting area to analyze here. Sun has some nice cloud computing and virtualization technologies that Oracle now owns. So Oracle could become a leading provider of cloud
computing offerings with this acquisition.
By acquiring Sun, Oracle also stops IBM from acquiring the company. Stopping a competitor is not usually the most compelling reason to acquire a company; but when there are as many synergies as this
one for Oracle, throwing up that roadblock for IBM must only have been viewed as additional gravy when they made the decision to pull the trigger.
And, indeed, this acquisition is very disruptive to the market, for a number of reasons. One fails to see how Oracle won’t take advantage of Open Office and Star Office (the productivity suites
it gains from Sun) to throw a wedge into Microsoft’s best selling Office productivity suite. Java developers will be wondering just what the heck Oracle will do now that it controls the Java
reigns. The non-IBM server big boys, HP and Dell, have got to be re-evaluating their long-term plans for partnering with Oracle.
What are the risks for Oracle in acquiring Sun? Oh, there are some. Oracle has no experience with hardware, but hardware and related sales account for 80% of Sun’s revenues. Oracle can mitigate
this risk by leveraging the experienced folks it acquires from Sun. And Oracle has always been savvy about adding executives with skills it lacks when it needs to. Oracle also must contend with the
deteriorating market share for Sun servers if it wishes to halt the erosion.
Another point of consideration is what will happen to Oracle’s relationship with HP. I’m sure the acquisition of Sun will strain that relationship, perhaps considerably. But both
companies will want the relationship to continue. Oracle has for years sold itself as a DBMS that works across all types of heterogeneous hardware. And HP really has nowhere else to go for database
market share.
And what of the relationship between SAP and Sun? Oh, that is dead, dead, dead!
The deal is quite costly – one of the biggest Oracle has taken on. But as Oracle points out in the press release, when Sun’s cash assets are taken into account, the deal is valued at $5.6
billion. Oracle expects Sun to add at least 15 cents to its income on a non-GAAP basis in the first full year after closing, and $1.5 billion to its non-GAAP operating profit… and $2 billion
in the second year. Even though Oracle is confident it can achieve these goals, actually doing so may prove to be difficult. Some analysts estimate that Oracle may need to cut up to 10,000 former Sun
employees (from a staff of 30,000) to make its profit targets. But how can Oracle do that and maintain the hardware expertise it needs to maintain Sun’s primary business?
So what will happen next? Well, Oracle could be on the lookout to bolster its services offerings if it is truly looking to become the next IBM; Microsoft will probably start to eyeball some
enterprise hardware and software companies (Cisco? EMC?); and HP and Dell will scurry around looking for ways to replace Oracle as a partner. I bet that Dell will try to marry itself closer to
Microsoft.
Oracle has now acquired 55 companies since 2005, and Sun Microsystems is the biggest. But Oracle’s users are intrigued. “The Sun acquisition is very exciting news for Oracle users,”
said Ian Abramson, president of the International Oracle User Group. I doubt that Sun’s customers are as “excited,” though.
One thing is for sure, the rest of this year could get very interesting indeed!
Oh, There Were Other Acquisitions in the Second Quarter, Too!
IBM did make an acquisition in the second quarter though, just not one the size of the Oracle/Sun deal. In early May, IBM announced that it had acquired Exeros, a private company that produces
Exeros Discovery X-Profiler. Financial details were not disclosed.
The Exeros technology helps organizations uncover hidden relationships between databases, helping users make sense of disparate data sources much faster than otherwise possible. The capabilities
provided by Exeros seem to match up well with IBM’s recent spate of acquisition in the data management space. Look for IBM to integrate Exeros into its Optim product line.
IBM claims that the acquisition of Exeros will further IBM’s Information on Demand initiative, which is aimed at enabling customers to use information as a competitive and strategic business asset.
Ambuj Goyal, general manager of IBM’s Information Management groups stated that “(t)he combination of IBM and Exeros will enable companies to more intelligently manage their data across
all formats and computing platforms, creating a smarter enterprise.” For more information about Information on Demand and IBM Information Management, click here.
Incidentally, Exeros Discovery X-Profiler is the product upon which CA’s ERwin Data Profiler is based. It will be interesting to see what happens to the CA offering as I’m guessing it will be
unlikely that IBM will want to continue that relationship.
And Oracle was not content to simply acquire Sun Microsystems, they announced yet another acquisition this past quarter, too. In mid-May the company acquired Virtual Iron Software, Inc., a provider
of server virtualization management software that enables dynamic resource and capacity management in virtualized data centers.
Virtualization management looms as one of the next big frontiers for server virtualization vendors. “Industry trends are driving demand for virtualization as a way to reduce operating expenses
and support green IT strategies without sacrificing quality of service,” said Wim Coekaerts, Oracle Vice President of Linux and Virtualization Engineering. “With the addition of Virtual
Iron, Oracle expects to enable customers to more dynamically manage their server capacity and optimize their power consumption.”
The combination of Oracle and Virtual Iron supports Oracle’s strategy to provide complete, full stack management across the virtual and physical enterprise and should help to provide customers
with comprehensive and dynamic virtualization management. Oracle intends the combined suite of products (Oracle’s VM coupled with Virtual Iron software) to simplify the deployment and
configuration of physical servers, virtual machines, and applications while providing a highly available platform for hosting Oracle software and other enterprise applications.
Oracle also specified that it intends to combine Virtual Iron technology with Oracle Enterprise Manager to enable customers to be more agile in meeting application service levels for virtual
environments.
And Yet the Layoffs Continue…
Regular readers of this column will recall the biggest DBMS news of the first quarter was the economy and its impact on the database software sector. Oracle, IBM and Microsoft all announced
layoffs in the first quarter. And the economic impact – most notably, more jobs being cut – continued in the second quarter of 2009.
Early in the quarter, details of internal IBM documents outlining layoffs were leaked. And the news was grim, at least if you were an IBM employee. Those internal documents indicated that IBM would
be eliminating more than a thousand positions (1,674 cuts) in its Global Business Services unit alone.
Information Week reported, “In an effort to rebalance skills, eliminate redundancies, and deliver greater economic efficiencies, IBM Global Business Services is announcing a resource action
affecting U.S. employees in Application Services,” says an IBM document dated March 26, 2009. Don’t you just love the term “resource action?” No? Neither do I! I can understand
that a company might have to lay people off, but at least treat them like humans and call it what it is – a layoff, not a “resource action.”
Evidently those resources impacted by this action are eligible to receive one week of pay for each six months worked, but
not more than 26 weeks. The company will also offer some healthcare and retraining benefits. Of course, axed employees will have to sign away their rights to sue the company. But that is standard
operating procedure for most layoffs when a severance package is offered.
Not everyone sees these layoffs merely as a reaction to the current economic conditions, though. Alliance@IBM, an IBM employee group, believes that IBM is eliminating U.S. jobs and moving the same
jobs to low-cost overseas locations, such as India. There was also a political reaction to IBM’s “resource action” and Alliance@IBM’s allegations. New York State Assembly
member Greg Ball, a Republican whose district includes IBM’s home county of Westchester, questioned how a company that receives taxpayer assistance can cut employees and move jobs overseas. Ball
called for a legislative hearing into the matter.
In July 2008, New York said that it would provide $140 million in grants to IBM, which in turn was investing $1.5 billion to create 1,000 new jobs in nanotechnology. The agreement also included $65
million in provisions to help IBM retain jobs at its East Fishkill plant in Dutchess County, an area also represented by Ball.
Of course, IBM was not the only company “in the news” this quarter for laying folks off. In early May, Microsoft continued with the job elimination it announced in the first quarter. The
company cut over a thousand positions (1,200 cuts), mostly at its headquarters in Redmond, WA.
Remember that in January 2009, Microsoft indicated that it would cut as many as 5,000 jobs by June 2010. About 1,400 jobs were cut worldwide in January. More interestingly, perhaps, is the memo sent
by Microsoft CEO Steve Ballmer to Microsoft employees on May 5th letting them know about the latest round of layoffs. The full text of the email can be read online at:http://www.businessweek.com/the_thread/techbeat/archives/2009/01/steve_ballmers.html.
So it looks like there will be more layoffs at Microsoft. At least that is what I would take away from this email if I were a Microsoft employee reading it at my desk. How about you?
How About Revenue at the Big Three?
Now, as we do each quarter, let’s turn our attention to the revenue announcements of the Big Three: Oracle, IBM and Microsoft.
In late March, Oracle announced its fiscal third quarter numbers, and they were not too bad given the current economic climate. The company reported a 1% decline in net income to $1.32 billion,
compared to $1.34 billion in the year-ago quarter. Revenue was up 2% to $5.45 billion. The decline was attributed to declining value of foreign currencies against the U.S. dollar.
Oracle’s operating income during the quarter came in at $1.94 billion, up 4%. And software revenue grew 5% to $4.4 billion, but new software license revenue was down 6% to $1.5 billion;
software license updates and support revenue was up 11% to $2.91 billion. And quarterly revenue from services dropped 8% to $1.02 billion, down from $1.1 billion from the same quarter last
year.
Oracle president Charles Phillips said: “If you look past the effect of exchange rates, our new software license revenues for this quarter were higher than our new software license revenues for Q3
of last year. Achieving constant currency growth in new software license sales in this very challenging economy shows that we continue to beat our competitors in both technology and
applications.”
And Oracle announced it would pay out its very first dividend, which is interesting since the recession has caused many organizations to eliminate dividends. Oracle will pay five cents per share,
starting in May 2009; this will cost the company about $1 billion per year.
The news at Microsoft was not nearly as “good,” where the company reported its worst third quarter earnings ever in late April. Revenues came in at $13.65 billion for the third quarter
ended March 31, 2009, which is a 6% decline from the same period last year. Operating income was $4.44 billion, an increase of 3%; net income was $2.98 billion, which is a drop of 32%; and, diluted
earnings per share for the quarter were $0.33 per share, which represented a decline of 30% (all compared with same period last period). This is quite a change from the third quarter of 2008 in which
Microsoft touted its “record third quarter revenue.”
This was Microsoft’s first ever year over year quarterly sales drop. Bad news, but actually quite impressive given that Microsoft has been publicly traded for 23 years! And keep in mind that
the financial results for the quarter included $290 million of severance charges related to Microsoft’s plan to eliminate up to 5,000 jobs. It also reflects $420 million of investment losses.
These two charges reduced earnings per 6 cents per share.
“While market conditions remained weak during the quarter, I was pleased with the organization’s ability to offset revenue pressures with the swift implementation of cost-savings
initiatives,” said Chris Liddell, chief financial officer at Microsoft. And the company also provided updated guidance for operating expenses ranging from $26.7 billion to $26.9 billion,
including severance charges, for the full year ending June 30, 2009.
In late April, IBM reported its fiscal first quarter earnings, which were basically flat, but reasonable in this economy. IBM reported diluted earnings of $1.70 per share compared with $1.64 per
share in the first quarter of 2008, which is an increase of 4%. But first quarter net income was $2.30 billion compared with $2.32 billion in the same quarter last year, which is a decrease of 1%.
Total revenues for the first quarter of 2009 came in at $21.7 billion, a decrease of 11% from the first quarter of 2008.
“IBM continued to perform well in a very difficult economic environment. This was due to our long-term strategic focus: shifting into software and services, divesting of commodity businesses,
and creating solutions that help clients reduce cost and conserve capital. At the same time, we have a disciplined approach to cost and expense management giving us a strong financial position,”
said IBM chairman, president and chief executive officer, Sam Palmisano.
IBM indicated that it is ahead of pace for its 2010 road map of $10 to $11 per share, and that it expects full year 2009 earnings of at least $9.20 per share.
Given the state of the economy then, it appears as if the Big Three DBMS vendors are performing reasonably well… or, at least, things could be a lot worse!
What’s New Brewing with DB2?
This quarter also saw the announcement of a new point release of DB2 – Version 9.7 – from IBM. The announcement of DB2 9.7, also known by the code name Cobra, was made in late April
and it offers some interesting new functionality.
Probably at the top of the new features list are the workload management capabilities. Among the new workload management features in DB2 9.7 is the ability to set different priorities for different
types of applications. DB2 users will also get the ability to set with conditional workloads so that can applications can change when a workload threshold is exceeded. Mainframe users will recognize
these capabilities because they’ve been available there for years with Workload Manager, but they are new to the Linux, Unix and Windows world. The bottom line is that DB2’s workload
management features can improve the performance of mission critical applications by better balancing system workload.
Additionally, IBM is claiming that improved autonomics and system administration capabilities can help DBAs better manage their growing data, leading to a reduction in the time it takes to complete
administrative tasks by up to 35%.
A particularly interesting new capability built into DB2 9.7 is its ability to simplify the migration of data from other DBMS platforms (well, Oracle mostly) to DB2. What IBM did here is license
technology from EnterpriseDB, the open source database vendor that makes its living off of mimicking Oracle. The capabilities in this licensed technology enable customers to run applications written
for Oracle Database even though they are using DB2 9.7.
Other new features include new user interface that simplifies administrative tasks, new technology for analyzing XML data, more effective data compression, more online schema operations, support for
weak data typing, new data types, new built-in functions, lots of SQL enhancements and more.
At the same time, IBM also announced a new release of its InfoSphere data warehousing technology. Both DB2 9.7 and the new release of InfoSphere were supposed to be generally available by the end of
June (which will be in the past when you read this, but is in the future as I type this)!
The IBM press release announcing DB2 9.7 is available online at: http://www.ibm.com/press/us/en/pressrelease/27279.wss.
Microsoft SQL Server 2008 Service Pack SP1
And IBM was not the only DBMS provider with a new “release” this quarter. In early April Microsoft unleashed the SP1 service pack which rolls up all the fixes since the initial
launch of SQL Server 2008.
Of course, this announcement was not as significant as the DB2 9.7 announcement, but it has the benefit of actually being released this quarter. And it could have a big impact on the uptake of SQL
Server 2008. Many SQL Server 2000 and 2005 shops have been waiting for SP1 before upgrading to SQL Server 2008.
Other than the rollup of fixes, SP1 also introduces at least one new feature called Slipstream, which should make it easier to deploy SQL Server 2008. Slipstream enables users to install the DBMS and
the service pack all at once.
This approach is basically a new one for Microsoft – delivering a service pack that consists almost exclusively of fixes, instead of a combination of fixes and new functionality. Microsoft is
hoping that this approach will deliver a more stable environment which will lure users to upgrade. Finally! In my opinion, new functionality should be delivered in new releases, not fix packs.
Amazon Cloud Database
In early April, Amazon announced its new cloud database offering. Amazon Web Services is probably the biggest player in the cloud computing market. Most of Amazon’s cloud business is for
its S3 storage service S3 and its EC2 computing power as a service offerings. Prior to April, Amazon also had a cloud-based database offering called SimpleDB. SimpleDB is a key-value pair database
that targets web developers who don’t need or want a relational database.
The new hosted service from Amazon is designed to simplify the use of the Hadoop implementation of the MapReduce programming model for processing large data sets in processor clusters. The offering
is called Amazon Elastic MapReduce, and it is aimed at developers whose applications need to process very large amounts of data.
Amazon’s new cloud database offering is a relational DBMS deployment delivered in partnership with IBM. As with other Amazon Web Services cloud offerings, Amazon charges for Elastic MapReduce
based on usage, without requiring any minimum fee.
It remains to be seen whether database implementations will grow and prosper in the cloud; but if they do, Amazon appears to be ideally suited to benefit. So this looks like a good move to me.
The Open Database Alliance
And in mid May, in a move not un-entirely unrelated to Oracle’s acquisition of Sun Microsystems, Monty Program Ab, a MySQL database engineering company, and Percona, a MySQL services and
support firm, announced “The Open Database Alliance.” Monty Program Ab was founded by Monty Widenius, the originator of the MySQL DBMS.
The alliance is promoted as a vendor-neutral consortium designed to become the industry hub for the MySQL open source DBMS. The alliance hopes to promote MySQL and derivative code, binaries,
training, support and other enhancements for the MySQL community.
The Open Database Alliance will work cooperatively to provide software, support and services for MariaDB, a community-developed branch of MySQL. More information about the alliance can be found on
the web at opendatabasealliance.com.
Summary
Thus ends this quarter’s edition of the Database Report. And it was a blockbuster of a quarter for the database industry. As we continue to fight an economic recession the database world
remains active, thriving, and entertaining. What remains for the rest of 2009? If you want to know, be sure to check in with The Database Report each quarter to review all the news that matters
regarding your favorite database companies and software.
