The Database Report – January 2001

Welcome to the first edition of a regular quarterly column I will be writing for called The Database Report. Regular readers of the column will be kept up-to-date
on the latest developments and technology coming out of the large DBMS vendors, namely Oracle, IBM, Microsoft, and to a lesser extent, Informix and Sybase. The last quarter of 2000 was indeed an
active one. Let’s start the database report off with an examination of the goings-on at Oracle Corporation.

An Update on Oracle

Oracle has received quite a bit of publicity lately. Some of the PR has been great. And while it is said that there is no such thing as bad publicity, some of the recent news has to be
disappointing to Oracle. Let’s start off on a positive note, though.

The first bit of important news is the announcement of Oracle9i. Oracle9i is the latest and greatest release of Oracle’s stalwart DBMS. Announced at Oracle Open World in November 2000, Oracle9i
provides a single-engine architecture with built-in ETL, OLAP, and real-time personalization. Additionally, Oracle is integrating more e-business capabilities into the DBMS such as portal services
and clickstream intelligence. There are a lot of buzzwords in those two sentences, but at the core of the strategy is integration and consolidation. Oracle is combining what once were independent
products into the Oracle9i database engine.

Consolidation of products into DBMS features is a continuing trend for database management systems. More and more functionality is being sucked into the DBMS. This is good for customers because it
simplifies the number of distinct product offerings and delivers more functionality in a single “product.” The question remains, though: “Just how integrated will the functionality be?” – the
better the integration, the better the end-result for the customer. But this integration will complicate database administration because the DBMS is more complex and therefore, more difficult to
learn, manage, and maintain. And not everyone will want to use all of these new features. The more features a product has the more difficult it is to learn – even if you are not using all of those
features. I remember when Microsoft Word was a lot easier to use than it is today, and I didn’t need a talking paper clip to figure out the technology.

Oracle’s overall strategy seems to consist of consolidation and integration of many of its standalone products into two distinct product lines: the DBMS and the application server. The benefit to
the customer is more functionality and potentially, less difficulty understanding what product is required to support specific requirements. The drawback for customers is that a single supplier (in
this case, Oracle) gains more control over your infrastructure. Instead of one vendor supplying the DBMS and another the data warehousing support (ETL and OLAP), you now have one vendor supplying
it all, Oracle. Many sites balk at having back-end and front-end features supplied by a single company because it places too much control over their business in the hands of a single supplier.

Oracle’s Management Team Problems

Their technology notwithstanding, Oracle is experiencing some trouble keeping its high-level executive management team intact. In late November, Oracle lost their second highest-ranking executive
for the second time in less than six months. This time it was Gary Bloom, who resigned from Oracle to become CEO of Veritas Software, a successful storage management software provider. Late in June
2000, Ray Lane, the former Oracle COO resigned to pursue a career in the world of venture capital.

Although the specific reasons for the departure of each executive seem to differ, the actual reason for both most likely can be attributed to the “Larry” factor. Lane was a highly sought after
executive and rumors abounded that he would leave Oracle for more lucrative or powerful positions throughout his tenure at Oracle. But speculation as to the real reason that Lane left Oracle was
that Larry Ellison wanted more control over the day-to-day operations of the company, thereby diminishing Lane’s power. Similarly, Bloom had CEO-level aspirations – a position to which it was
unlikely he would ever ascend at Oracle as long as Ellison was around. So after 14 years with Oracle, Bloom resigned in late November 2000 to accept a position as CEO at Veritas.

Although the management exodus at Oracle is cause for concern, the loss of any single individual will not damage a company the size and stature of Oracle Corp. Of course, that would not include the
loss of Ellison, whose departure would be quite damaging. Larry is Oracle’s visionary and is the key to Oracle’s current success. But he needs quality management to oversee a company the size
Oracle. But quality management may be difficult for Oracle to attract as long as Ellison is in charge. It seems that Ellison wants to hold the reigns of the company very tightly. Any future
management hire may find that he or she has limited power.

In more positive employment new for Oracle, in late November former White House Press Secretary Joe Lockhart joined Oracle’s senior management team, reporting directly to Chairman and CEO Larry
Ellison. Lockhart’s first task will be to refine and communicate Oracle’s business strategy. Larry Ellison said that “Joe brings perspective and experience not commonly found in Silicon Valley
companies.” That is true. But Larry presents public relations issues not experienced by other Silicon Valley companies – so it is a match made in heaven. The addition of Lockhart does not resolve
the issue of who is second-in-command at Oracle, but it should help to strengthen Oracle’s (and Larry’s) public image.

A Bad Support Rap

Speaking of Oracle’s public image, a recent survey conducted at the annual UK Oracle User Group conference in Birmingham is critical of Oracle’s technical support. The survey completed by 191
Oracle users rated the company on support and maintenance. The results: one in three Oracle users has difficulty maintaining their software and one in five is unhappy with Oracle’s level of

Software support is increasingly important to customers because database products are becoming more and more complex. Keeping the DBMS up-to-date with fixes and patches is a difficult problem even
with excellent support. Oracle needs to respond to the consistent criticism of their software support, or risk losing business to a company with a better reputation for support, like IBM.

The Million Dollar Guarantee

On another, somewhat ridiculous front, Oracle is offering to pay customers $1 million dollars if they cannot triple the performance of their web site by switching to an Oracle8i DBMS and an
Oracle9i application server. Looks intriguing? Look again.

The first problem is the cost of the solution. To participate in the challenge the “customer” is required to pay for all software and services – even if Oracle fails to deliver the three-fold
performance. And Oracle requires that they be given 90 days to provide tuning for the site. That sounds reasonable, but the client is required to pay for the tuning services. Well, at this point
the customer may have paid much more than $1 million already!

But there is another problem. This is your web site for heaven’s sake! By converting to another database and application server the customer risks an outage to convert. And how much business will
be lost during the outage required to convert from your current architecture to the new Oracle architecture.

Finally, your company is forbidden to talk about the experience if Oracle fails, but if successful the organization must agree to be used in Oracle advertising campaigns. So, this $1 million
challenge is an interesting marketing tactic, but the details will make the program look unappetizing to most organizations.

Oracle’s Financials

And the final bit of Oracle news involves their financial condition. Oracle announced its second quarter earnings on December 14, 2000, and handily beat consensus analyst expectations. Profits rose
from $384 million in 2Q1999 to $623 million, a 62% improvement. And revenues rose by $400,000 to $2.7 billion.

Sales of Oracle’s database server software increased to $775 million, a 19% increase. This result fell within analyst predictions that ranged from a low of 18% to a high of 22%. Sales of Oracle’s
applications rose a healthy 66% to $279 million, well ahead of expectations of that ranged between 48% to 58%.

For the long term, Oracle appears to be on track with its strategy and development plans. They just need to continue meeting or exceeding revenue expectations and overcome some of their recent HR
and support problems.

And at IBM…

IBM was busy during the latter half of 2000, too. In September IBM announced their entry into the database tools market. Well, entry is probably the wrong word to use here. IBM has offered database
tools for DB2 for a while now. But those offerings have been limited in number. The interesting thing about this is that IBM is telling their customers that by selling them additional management
tools they will reduce the customer’s TCO for DB2. This is hard to believe. First of all, they are going to start charging for base utilities like LOAD and REORG that were previously free.
Secondly, they are going to begin selling tools that compete with products from ISVs like BMC Software and Computer Associates. At least initially though, IBM will not have replacement technology
for all of the ISV offerings. IBM’s approach is to charge less than the ISVs for similar technology. But will a cheaper product that does less actually reduce TCO? Probably not.

And the replacement technology IBM will be quite inferior from a functionality viewpoint to the ISV offerings. So, thinking logically for a minute, wouldn’t reduced functionality translate into a
higher TCO as IBM’s customers need to manually replace the functionality gap between the IBM offerings and the ISV offerings? Clearly IBM has some challenges ahead of them in the database tools

Support and Maintenance

A much more positive note for IBM is provided in customer’s perception of their support and maintenance. Especially in light of the recent bad news Oracle received on this front.

In December 2000, InformationWeek published a survey in which customers were asked to rank IBM’s DB2, Oracle, Microsoft SQL Server, Informix and Sybase in ten categories. In short, IBM and DB2
blew away the competition. IBM and DB2 ranked first in the following categories: customer referrals, service and responsiveness, industry expertise, reliability and availability, scalability and
VLDB, service-level agreements, and strategic advice. In the only categories IBM did not win: Oracle won for features and innovation (IBM was second), and Microsoft won for pricing and value (IBM
was third) and programming expertise (IBM was second).

And IBM’s ranking is even more impressive when viewed in context with the remainder of the survey. Especially when the fact that customer service ranked second just behind reliability and
availability as most important criteria for selecting a DBMS vendor. Clearly, customers enjoy their relationship with IBM, at least as it pertains to their usage of DB2.

Details of the survey can be found at


Well, that is about all we have room to discuss in this edition of the database report. But be sure to check out this column in future issues of for more news and analysis of the DBMS


submit to reddit

About Craig Mullins

Craig S. Mullins is a data management strategist and principal consultant for Mullins Consulting, Inc. He has three decades of experience in the field of database management, including working with DB2 for z/OS since Version 1. Craig is also an IBM Information Champion and is the author of two books: DB2 Developer’s Guide and Database Administration:The Complete Guide to Practices and Procedures. You can contact Craig via his website.