Even though each of the major DBMS vendors made news during the third quarter of 2003, this past quarter was significantly dominated by Oracle. The three biggest DBMS-related news stories of the
quarter were all dominated by Oracle: the annual Oracle World conference, a major new release of Oracle, and the on-going saga of Oracle versus Peoplesoft. We’ll take a look at each of these
events, and more, in this edition of The Database Report.
Oracle10g: The Nitty Griddy
The biggest news of the past quarter is the announcement of the latest version of Oracle’s DBMS named Oracle 10g. This new version is expected to ship before the end of this year (that is, 2003).
The “g” stands for grid and Oracle 10g is so named because it is built around the concept of enterprise grid computing. The basic idea behind grid computing is to ease the development of new
systems by basing them on easily assembled pools of standard components such as storage and servers.
Grid computing has its roots in highly parallel computing where complex problems required the capacity and capabilities of multiple computers to solve. Grid computing offers the promise of reduced
cost because smaller, cheaper computing systems can be coupled in a grid to achieve solutions comparable to what can be accomplished on larger, more expensive computing systems. Furthermore, grids
have the potential to take advantage of under-utilized computing power that otherwise would remain unused.
More details about Oracle and grid computing can be found in Oracle’s white paper titled “Oracle and the Grid,” which can be found on Oracle’s web site at the following location: http://otn.oracle.com/products/oracle91/grid_computing/OracleGripWP.pdf.
The overall goal of grid computing is to reduce costs while delivering truly scalable systems that provide reasonable performance. Let’s examine some of the technology and new features that enable
grid computing in Oracle 10g.
Automatic Storage Management, or ASM, helps to simplify the process of laying out data files across disk storage devices for Oracle databases. With ASM the need for a file system is eliminated and
Oracle 10g can be used to directly manage raw disks. ASM can automatically stripe database data across all disks – all with limited or no management requirements. With ASM comes the ability to
perform optional mirroring and the ability to add and drop disks online. ASM also can automatically detect hot spots and rebalance data to eliminate these problem areas. All of these features
improve data availability, which is a major concern in this day and age of 24 by 7 and web-based applications.
Oracle 10g expands the ability to store and manage very large databases – or what Oracle is calling Ultra Large Databases (ULDBs). Databases can be created in the exabytes range (that is, millions
of terabytes). Such support is provided with ultra large files that can store terabytes in a single file and terabyte-plus LOB columns, coupled with ASM to manage these ULDBs.
Data movement is improved because data will need to be moved through the grid at high speeds. Oracle Data Pump is a new feature that offers high-speed movement of data and metadata. Bulk data can
be moved in parallel across multiple platforms and database versions using the Oracle Data Pump.
Cluster management for Oracle Real Application Clusters (RAC) is improved with Automatic Service Provisioning, or ASP. With ASP servers can be automatically allocated to workloads. This makes it
possible for clients to be automatically routed to the server with the lowest current workload for processing. RAC and ASP work in conjunction to facilitate large workloads to be processed across
multiple, connected Oracle servers.
Oracle 10g offers enhanced flashback capability. With flashback the DBA can basically rewind database activity to perform point-in-time recovery without requiring a restore and then rolling forward
through the log.
Additionally, Oracle 10g offers many other ease-of-use features such as:
- Improvements to Oracle Enterprise Manager that enables grid control;
- The ability to reduce downtime with rolling upgrades of hardware, operating systems, and DBMS versions;
- An improved PL/SQL offering much faster performance;
- Online redefinition of tables and indexes;
- Improved security features with Oracle Identity Management to deliver fine grained audit, generic auditing, virtual private database, and Oracle label security features;
- Enhanced data warehousing and business intelligence capabilities including improved OLAP analytics, a data mining GUI, and a new SQL model feature;
- Support for XML within Oracle databases has been improved significantly; and
- Oracle claims that 10g will be quicker to install than previous versions.
Oracle also announced Oracle Application Server 10g, which makes applications available over the grid.
The bottom line here is that Oracle 10g is an exciting and interesting new version of Oracle’s DBMS. It offers many new features with new functionality that should improve the manageability and
cost structure for your Oracle systems. Speaking of Price
Oracle is touting the grid capabilities of Oracle 10g as cost saving because systems can be more uniform in nature. Many Oracle customers run multiple versions of the DBMS – and that can drive up
the cost of ownership for Oracle sites. But with grid computing it becomes easier to manage a single image and then clone that image to other servers.
But, somewhat lost amidst the grid computing announcements at the Oracle World show (held the first week of September 2003) were some pricing changes that Oracle pre-announced. Although the exact
details were not announced, Larry Ellison told the Oracle World attendees that Oracle would be publishing new pricing for its software.
The basics of the pricing action seem to be that Oracle will reduce the price of RAC – a move that should help to speed its adoption at customer sites. Oracle is intent on spreading the RAC gospel
to the masses and lowering the cost of entry for RAC will help.
Oracle needs to improve its pricing to be successful in the SMB market. Microsoft is known for its success in this market and even IBM has been pricing its DB2 software more attractively for the
SMB market with its recent DB2 Express offering. So, Oracle seems to have accepted the challenge and will be lowering its prices. But it will take more than just lowering list prices for customers
to start viewing Oracle as a “friendly” company to do business with.
Speaking of Oracle World
Usually the only news coming out of Oracle’s conference held every year in San Francisco is technology and marketing related. But not this year! Immediately following an afternoon keynote address
the site of the conference – the Moscone Center – had to be evacuated due to a bomb threat. Now I’ve heard of speeches that bomb but this one must have taken the cake!
Seriously, though, with over 15,000 attendees to account for, this was not a simple exercise. Fortunately, no bomb was found, there were no explosions, and no one was harmed.
ORCL – PSFT: An Acquisition in Limbo?
So what about Oracle’s impending PeopleSoft acquisition?
Early in the quarter, right before the July 4th American holiday, PeopleSoft announced its most recent quarterly earnings – and they were better than most expected. This was bad news for Oracle
because a healthier PeopleSoft is likely to be a more difficult company to acquire. If this recent good news is a harbinger of continuing revenue growth for PeopleSoft then its stock will likely
climb causing Oracle’s bid for the company to become less enticing to PeopleSoft stockholders. Of course, Oracle has already raised the bid amount once, and they may indeed have to raise it again.
PeopleSoft’s earnings were in the $490 to $500 million range, well ahead of the analyst expectations, which were set at about $443 million for the quarter. Of course, Craig Conway, CEO of
PeopleSoft, used this good fortune to bash Oracle. He intimated that the robust quarterly earnings were an indicator that his customers were “determined not to let Oracle’s hostile takeover
attempt hurt PeopleSoft.” Well, perhaps, but I truly doubt that is the entire case. Quarterly earnings do not usually turn on a dime like that and the Oracle takeover bid came not at the beginning
of last quarter, but near the end. Actually, the increased earnings are a good sign for PeopleSoft – it shows that customers are still happy with the software and are increasing their buying. And
it shows that PeopleSoft sales staff was able to continue signing contracts even with the distraction of a possible impending takeover looming over them.
So, what does this actually mean for the outstanding Oracle takeover bid? Well, part of the Oracle argument for acquiring PeopleSoft was that PeopleSoft was having trouble maintaining and growing
its customer base. Indeed, Larry Ellison’s position seems to be that PeopleSoft share holders should be bowing down and thanking Oracle for coming to their rescue. But with PeopleSoft’s improved
revenues this line of thinking is seriously flawed. Of course, Oracle can continue to point to PeopleSoft’s previous five quarters of declining results. Oracle will position this most recent
quarter as an aberration and continue to assert that PeopleSoft is a business in decline. Stay tuned next quarter to see if PeopleSoft can continue to grow earnings – or if it goes back into a
declining pattern.
PeopleSoft’s earnings notwithstanding, Oracle continued to extend its offer for PeopleSoft and remains “fully committed” to buying the rival company at its $7.3 billion bid. At the same time
PeopleSoft continues to pursue legal angles to thwart the acquisition touting Oracle’s use of unfair trade practices.
The most interesting complaint made by Peoplesoft is that Oracle is more interested in hurting PeopleSoft’s business than in acquiring the company. This claim gained some credibility when internal
Oracle e-mails were released with some potentially damaging content. Several e-mails from Oracle executives stated things like “We’ve certainly wounded (PeopleSoft)…” and “even if we don’t
end up closing the deal this is going to take PeopleSoft some time to recover…” Now does this sound more like the words of a company trying to damage PeopleSoft or acquire PeopleSoft? Of course,
Oracle disputes PeopleSoft’s allegations.
Meanwhile, Oracle sits on a vast reservoir of capital – $6.52 billion in cash. That money seems to be burning a hole in their pocket. If Oracle doesn’t end up with PeopleSoft then I’m sure Larry
will spend his money somewhere else.
My take on this matter remains the same: I believe that the Oracle bid will likely fail. PeopleSoft appears ready and willing to fight the takeover bid and such hostile takeovers are difficult to
pull off. The takeover also will have to clear US Department of Justice hurdles because it would diminish the competition in the ERP application software space. Of course, the US is not the only
jurisdiction in which Oracle’s hostile bid for PeopleSoft is being reviewed. Canadian antitrust regulators are also investigating the acquisition bid and that adds another barrier in the way of
this takeover attempt.
And PeopleSoft’s users are likely to be extremely vocal during these anti-trust hearings as they raise the specter of Oracle forcing PeopleSoft users to switch to Oracle databases. That is the
underlying story of this whole saga in my opinion. Oracle’s “bread and butter” is still its database business – not its application business. PeopleSoft gives its users the option of choosing
between several DBMS products, all of which are Oracle’s competitors. If Oracle owns the PeopleSoft code how long do you think they will continue to offer versions that run using competitive
databases? Larry Ellison is not exactly Bill Gates’ biggest fan, nor is he an advocate of IBM’s software. If Oracle can emerge from the anti-trust with a favorable ruling then Oracle will pursue
its outstanding lawsuit trying to force PeopleSoft to remove its ‘poison pill’ takeover defense. On the other hand, if the anti-trust courts rule against them, then Oracle will most likely drop
the PeopleSoft acquisition bid and go trolling for other acquisitions.
Other Acquisition Items of Interest
During the past quarter PeopleSoft also closed on its acquisition of J.D. Edwards, another ERP application software vendor. Oracle was originally against this acquisition, but has since changed its
stance and continues to pursue acquiring the combined PeopleSoft and J.D. Edwards.
Early in September 2003 Oracle conducted a “town hall” meeting to discuss their acquisition bid for PeopleSoft. The approach taken by Oracle at this event was both novel and compelling. In 1994
Oracle acquired the RDB DBMS from Digital Equipment Corporation. At this meeting Oracle deployed a customer of RDB to attest that Oracle has kept this “competing” DBMS technology maintained and
supported for almost a decade. Oracle then claimed that their intention is to treat the PeopleSoft acquisition the same way as they treated the RDB acquisition.
On the surface, this argument is somewhat compelling. It shows that Oracle has a history of supporting acquired technology that competes with its base technology. But there is a missing nuance
here. Can Oracle claim to continue to support PeopleSoft code while at the same time eliminating support for using a competing DBMS? I think so, and I think this will be Oracle’s medium term plan
for the PeopleSoft products. By continuing to support the PeopleSoft applications Oracle can claim to be keeping the products alive for PeopleSoft customers. But it is truly hard to imagine Oracle
continuing to allow users of PeopleSoft applications to run on DB2 or SQL Server. So if this acquisition does somehow occur, PeopleSoft users should start learning the Oracle DBMS sooner rather
than later. And IBM and Microsoft should plan on losing some customers to Oracle – namely, those customers using PeopleSoft applications. But let me reiterate, I still do not believe that this
acquisition will ultimately succeed.
And finally, some good news for Oracle: a small group of PeopleSoft share holders have filed class action lawsuits claiming that PeopleSoft should not have summarily rejected Oracle’s takeover
proposition. Eight PeopleSoft shareholders are currently pursuing separate class-action litigation against the company, alleging that PeopleSoft executives breached their fiduciary duties in
connection with their response to Oracle’s tender offer to purchase PeopleSoft.
This ongoing Oracle / PeopleSoft courtship is truly becoming quite the goat rodeo, isn’t it? It is rather strange but kind of fun to watch – and you never know what those goats are going to do
next!
And in Other News From Redwood Shores …
It was, indeed, a very busy quarter for Oracle Corporation. Early in the quarter Oracle announced that it would be stepping up its support for the Mac platform by supporting the Mac OS X operating
system for its next generation DBMS technology. Oracle’s RAC clustering capabilities align nicely with Apple’s Xserve rack-mounted server systems.
Finally, let’s take a quick look at Oracle’s fiscal first quarter earnings. The company reported in mid-September that net income increased 28 percent as compared to the same quarter last year.
With earnings of approximately $440 million this translates to $0.08 per share. The results met analyst expectations.
Revenue for the quarter came in at $2.07 billion, up 2 percent over revenues of $2.03 billion a year earlier. However, analyst expectations for revenue were $2.14 billion, so Oracle missed that
target. But software firms are notorious for having slow first quarters, so the miss is not necessarily a harbinger of bad news for the future.
All is not Quiet on the Northwestern Front
Meanwhile, up in Redmond, Washington, Microsoft has definitely pushed back the Yukon release of SQL Server; Microsoft claims to be taking this action to allow the new version of SQL Server to reach
the market at the same time as an upcoming upgrade to Microsoft’s development tools. The major new features of Yukon will focus on improved availability and manageability. But I doubt the real
name of the new release will be SQL Server “g”.
As we reported in the last edition of The Database Report, the Yukon release of SQL Server was originally scheduled for general availability during the first half of 2004. The new target for
release is now the second half of 2004. What a difference six months can make, huh?
In other SQL Server news, Microsoft announced new Microsoft Certified Database Administrator (MCDBA) exams. Basically, the new Windows 2003 exams will be able to be used to fulfill the core
networking requirements of the MCDBA for SQL Server 2000. So, DBAs will be able to satisfy the networking requirements of MCDBS by taking one of the following exams:
- 70-290: Managing and Maintaining a Windows Server 2003 Environment (new)
- 70-291: Implementing, Managing, and Maintaining a Windows Server 2003 Network Infrastructure (new)
- 70-215: Windows 2000 Server (available since June 2000)
Microsoft has also added an elective exam, 70-293: Planning and Maintaining a Windows Server 2003 Network Infrastructure, to its roster of MCDBA exams.
Over at IBM
Early in the quarter IBM unleashed a new version of the Informix DBMS – Informix Dynamic Server (IDS) version 9.4. Along with claims of faster processing, IDS makes high availability data
replication and enterprise replication fully interoperable.
The new version also comes with simplified configuration and administration for large database systems and eliminates some limitations allowing larger databases (up to 128 petabytes) to be
supported.
So, IBM is making good on its promise to continue supporting the Informix code base – even as it loudly markets DB2 and proclaiming it to be the best relational DBMS on the market. Evidently IBM is
committed to rolling out new versions of the Informix product line approximately every 18 months or so.
Of course, the longer IBM supports and upgrades the Informix DBMS the longer IBM can continue to encourage Informix users to stick around. Some studies show that when Informix users convert to
another DBMS they are more inclined to choose Oracle than DB2. So continuing to support Informix can help IBM to save some customers. Of course, IBM continues to offer a transition toolkit to
facilitate an easy migration to DB2 from Informix for customers choosing to do so.
Also at IBM, the Arrowhead initiative is still on. Arrowhead is IBM’s ambitious project for merging the best features of the Informix product lines (IDS and XPS), the Red Brick warehouse and its
classic DB2 technologies. We’ll have to keep our eyes open for future delivery on this initiative.
Late in September 2003 IBM released a patch for a security vulnerability in DB2 Version 7.2. The problem, if uncorrected, allows malicious code to be run on DB2 systems using the permissions of a
root account. This is basically a buffer-overflow problem, the likes of which have been seen before in DBMS products (e.g. SQL Server). This problem is not quite as severe, though. It would take a
very skilled hacker with knowledge of both DB2 and Linux to exploit the vulnerability. Now, let’s hope that IBM can encourage everyone to patch their systems before the hackers can take advantage
of the loophole.
In further IBM news, let’s examine IBM’s recent quarterly earnings. IBM’s revenue came in at $21.6 billion, which was up 10 percent (3 percent at constant currency). IBM executives claimed that
they believe the company can and will do better as 2H03 evolves.
IBM’s software business recorded revenue of $3.5 billion constituting a 6 percent growth, but at constant currency a 2 percent decline. Even so, both the WebSphere and DB2 product lines registered
growth. IBM’s software lines include Tivoli for systems management, WebSphere for application servers, DB2 for data management, Lotus for PC software, and Rational for application development.
Results for each product line follow:
- Tivoli was flat at constant currency
- WebSphere was up 6 percent at constant currency
- DB2 was up 7 percent at constant currency
- Lotus down 12 percent at constant currency
- Rational was up 6 percent at constant currency
Additionally, IBM reported that DB2 software on the host environment was up, but performance on distributed platforms was particularly strong, up14 percent at constant currency.
But keep in mind; software is only IBM’s third largest business. Hardware is number one and hardware revenue was down 6 percent at constant currency. Better news was announced for IBM’s second
largest business – consulting. IBM Global Services was up 14 percent at constant currency.
Overall, IBM’s earnings appear to be a mixed bag – some good news, some bad. IBM has a good mix of businesses but the quality of their earnings growth appears to be slipping overall. Subsequent
quarters should shine light on these trends.
Ice, Ice Baby
IBM debuted a new DB2 offering at the annual Linux World conference named DB2 ICE. The ICE stands for Integrated Cluster Environment. OK, but what is it and why should you card?
Well, if you are interested in Linux, ICE is nice. DB2 ICE runs on the IBM eServer 325 (also announced at the same time) and IBM claims it can scale 125 times higher than any competitor. And with
the capacity to scale to as many as 1000 server nodes, the claim has credibility. If you run a 1000-node cluster you will be able to manage 16 exabytes of data using DB2 ICE.
So, Linux appears to be ready to table your enterprise level high availability and performance computing needs. Open Source DBMS News
In other open source news MySQL and SAP DB have agreed to merge. Well, more accurately, MySQL has agreed to acquire the open source DBMS assets of SAP – SAP DB. Evidently the SAP DB DBMS will be
renamed and provided under the General Public License (GPL) of openly available software.
This is probably a good move for both companies. SAP gets to unload a database technology that is not in line with its core ERP competencies and which few SAP customers are actually using. MySQL
gets a closer partnership with SAP and a more open avenue to explore working its way into SAP’s large corporate customer base.
But the bottom line here is that open source software is still on the fringes of the DBMS world. We may be willing to try an open source operating system like Linux, but rare is the customer who is
using open source database technology to build large, enterprise level databases.
Sybase Woes
The Big Three of IBM, Oracle, and Microsoft are not the only newsmakers in the DBMS space this quarter. Sybase warned that it would take an additional restructuring hit of $2.1 million in its
current quarter. This adds to the ongoing cost of restructuring with the total now spent this year alone being $11.5 million. Sybase laid off 215 employees in its second quarter and planned to cut
an additional 25 this quarter.
The financial woes at Sybase further clarify the Big Three DBMS vendors as IBM, Oracle, and Microsoft. Sybase was a major player several years ago, but they just cannot compete with the big boys on
their turf any longer. Sybase saw its profits fall by 30% to $14.6 million in the second quarter (ended June 30, 2003) and even though this result was higher than analyst expectations it showcases
an ongoing downward trend. Sybase is in the process of adopting a strategy that focuses on mobile software, applications integration and key verticals. With such an approach Sybase may carve out a
comfortable niche in the database world – but they’re never going to challenge the Big Three again.
And Another Quarter Bites the Dust
And so ends the third quarter of 2003. Amid the bluster of acquisition news some major announcements about database technology were actually made. But the DBMS market is an always interesting
market and who knows what the fourth quarter will hold for us? Well, you will, if you tune in to TDAN.com again next quarter for the next installment of The Database
Report.
