The rumors of Oracle jumping into open-source services have heated up again, this time fueled by analyst comments that caused Red Hat shares to drop $1.59 on Friday. Red Hat shares finished the day at $19.90, which was a slight improvement over its brief fall below its 52-week low of $19.71.
According to what Jefferies & Co. analyst Katherine Egbert wrote on Friday, "Our independent checks in the past two weeks indicate that Oracle seems to be close to introducing its own software 'stack.'"
Jefferies cut its price target on Red Hat from $24 to $21.
I might have to eat my hat come Oracle OpenWorld next week, but as far as Oracle spending money to pick up the engineers from an affordable open-source distribution goes--Mandriva or Ubuntu, say--I lean toward agreeing with those who say that Oracle buying anything makes no sense when it could just roll its own.
Why, then, did Larry get everybody all excited back in April when he said in his Financial Times interview that Oracle had been pondering acquiring an open-source vendor such as Red Hat or Novell in order to control its own stack, operating system and all?
Because he's Larry Ellison, and he likes to say things that make people jump.
The key point he made, and then Tom Kurian and Chuck Phillips made after that, is that Oracle doesn't want to go broke doing the stack thing.
As for Red Hat's stock dip? Well, it's been limping along, anyway. At the end of September the company reported quarterly profits that dipped 34 percent from a year earlier, thanks to a sharp increase in operating costs.
Whatever Oracle plans to announce next week, if anything, I can't see it having financial consequences for the big Linux services companies. I can't see Oracle spending money on this.
It will be pretty cool if I'm wrong, though.
Technology Business Research came out with some interesting thoughts on Oracle’s buy of data integration player Sunopsis.
The main takeaway: Oracle’s acting to protect the pre-eminence of its database. As it stands, the Redwood Shore-ites are planning to incorporate Sunopsis’ integration technology into Oracle Fusion Middleware, along with the company’s SOA, BI and Master Data Management tools.
As far as the database goes, Sunopsis’ ETL products will offer support for both Oracle and non-Oracle data sources, thereby easing the integration of legacy databases.
The removal of the legacy database sticking point will make application sales a lot easier, TBR points out. “Technology Business Research believes Oracle’s acquisition of … Sunopsis SA serves to protect the central role of the database and to reduce possible barriers to applications sales by easing the integration of legacy databases,” analyst Stuart Williams wrote in a note. “The acquisition improves Oracle’s capabilities to access and interoperate in heterogeneous database environments (such as those from IBM, Sybase, Teradata).”
With Sunopsis technology allowing transformations to happen inside the source or target database, rather than in an external transformation server, Oracle databases stay at the center of data processing, rather than as a peripheral service, Williams said. What that means, in theory, is that we’ll get closer to Oracle’s touted “one version of the truth,” with less risky, easier and simpler integration of multiple data sources.
“The distinction offered by Sunopsis of ‘ELT’ rather than ‘ETL’ technology will help to sustain Oracle’s primacy as the database market leader as well as help to sustain the strong margins on the database products,” Williams said. “Database or middleware sales generated over $7.9 billion in the company’s fiscal 2006, representing over 70 percent of the company’s software revenue. TBR believes the acquisition may help to sustain growth in Oracle core enterprise database market, which we estimate has experienced modest single-digit growth for several years.”
What that means for customers already using central data warehouses, data marts or Oracle’s master data management technology is that they can look forward to getting some pretty groovy new functionality. Those who operate heterogeneous data environments—i.e., most—now have an alternative to external ETL servers. Those hankering after new apps but who are tied to legacy databases can look forward to easier integration, not only between the legacy database and the new app but also with BI, MDM and data warehousing.
News isn’t so good for competitors, Williams says. The acquisition strikes down the problem of integration with non-Oracle databases, but integration is still a “significant barrier” for IBM, Sybase, Teradata, MySQL and Microsoft, he says.
This hits at the same space—i.e., data integration--IBM was looking at when it acquired Ascential and its ETL technology.
Williams thinks that with Sunopsis technology, Oracle may surpass the IBM/Ascential pairing, thus relegating DB2 to “a contributing rather than central role.”
In light of the acquisition, here are some thoughts Williams suggest we keep in mind when we interact with apps vendors going forward:
“Does the data always flow into a central Oracle database or does the middleware work equally well in integrating a single Oracle database into an otherwise non-Oracle IT environment? While both companies may interoperate below the application layer, SAP certainly does not want SAP application sales to generate database sales for Oracle. Look for database competitors to counter this technology development with ‘ELT’ functionality in their database products or middleware,” he writes.
It’s been long enough, and I’m ready to say the words: Oracle Database 11g.
When’s it coming? Sources tell me it’s due sometime in 2007. Don Burleson did this crazy little trend graph based on previous major releases and the time that elapses between them to come up with the average number of years between releases being 3.2. That makes it April 2007 for 11g, but I agree with Don in thinking the R-Shores crowd would aim to make a splash at the fall Oracle OpenWorld 2007.
Mark Rittman, in his BI/reporting/data warehousing blog, said he thinks we’ll start to hear news late this year about what 11g is going to be about:
“surely the first bits of news [will start] to emerge in late 2006 about the next, 11g release of the Oracle Database.”
So what’s in store in 11g? Sources say Oracle's already done a great deal of work in these areas:
Security: Big things include support for security policies and separation of security administration from system or database administration.
Storage: Oracle's done work to take out some of the sting of storage costs, including a new file system that does cool things like de-duplication and automatic compression/encryption.
An example: Today the ASM piece that comes free with the database lets you replace the need for a volume manager and can even mimic high-end SAN storage capability on cheap disks.
This will help those who want to store content, including stuff like Word documents, spreadsheets and XML documents.
That sounds familiar, doesn’t it? It’s the tune IBM sang in its DB2 9 marketing: content-content-content, XML-XML-XML.
Is Oracle reacting to IBM’s content/marketing push? Maybe mostly at the perception level, since it's had XML DB since 9i R1 (PDF).
Which also brings up some interesting diss-age that’s been put in my ear about DB2 9 and the market’s utter indifference to XML in databases. I’m not spreading the dissing until after I talk to IBM and other people in the know, but if you have thoughts on it, please drop me a line.
And 11g: I’d love to hear what you all want to see in it, along with your thoughts on the security and storage tweakage.
With a bulging work force of 56,000, 22 acquisitions over two years and recent financial results so good Wall Street blinked in surprise, Oracle on July 18 held a 2-hour-long self-love-fest for investors and financial analysts.
The Webcasted luncheon was light on news, with the exception of Oracle President Charles Phillips saying that there's just nobody left that's big enough to swallow. "Like we said in the recent past, we continue to review ideas as they surface," Phillips said. "But most [companies] out there, there's just not enough out there that's very large."
But while the meeting held scant news, it served as a perfect raspberry to archcompetitor SAP. SAP's earnings on July 13 revealed a financial bellyflop, with the German software maker missing all financial analysts' estimates in a Reuters poll of 24 analysts.
SAP missed expectations for license sales, operating earnings and total sales in the second quarter, results that sent shares down by up to 10 percent on July 14. Its new-software license revenue rose 8 percent, to 621 millon euros.
As Phillips pointed out in Oracle's luncheon, and as Oracle bragged about in a recent Wall Street Journal ad, SAP's growth is anemic compared with Oracle's when it comes to business applications. Oracle's most recent earnings showed it had grown 83 percent year over year and 31 percent in that quarter.
Phillips quoted research that showed that only 6 percent of SAP customers are running SAP's most recent product, MySAP. He contrasted that with Oracle customers, 94 percent of which are running the most recent version of Oracle e-Business Suite, he said.
"What that means is 1) users upgraded to our product, and 2) we don't charge for upgrading," he said. "Not only are we shipping additional functionality, but we can prove you can actually [afford to] upgrade to it and get functionality out of it."
Phillips laid out a number of areas in which Oracle beats SAP: industry expertise, standards, real-time analytics, relationships with integrators and incremental growth in the midmarket.
As far as industry expertise goes, Oracle gets "unique intellectual property, unique to the industry that's driving this business," with its acquisitions, including its retail expertise picked up with its buys of ProfitLogic and Retek.
On standards, Phillips said Oracle has "embraced" them, as opposed to SAP's hewing to proprietary techniques.
Phillips also said Oracle's relationship with integrators has "changed dramatically over the last couple of years."
"Going into the midmarket, we have so many more customers, so many more products," he said. "We're working together in a way we probably didn't two to three years ago."
Lastly, Oracle's making incremental gains in the midmarket. "We've always been big relative to SAP," he said. "There's no question our product is easier to use. And [SAP has] multiple products. Our product you can use in a small company as you grow into a large company."
Earlier in the month, when IBM announced the ship date for Viper/DB2 9, I got on the phone with Phillip Howard, an analyst at Bloor Research.
Beyond the hubbub over the dual XML/SQL engines in DB2 9, Howard told me he was impressed by IBM's Data Warehousing BCU (Balanced Configuration Unit)--a set of best practices, multidimensional clustering, summary tables and more to get DB2 preinstalled in a data warehousing environment quickly and without getting bogged down by on-site configuration issues.
Howard opined that IBM was the only data warehousing vendor to be taking proactive steps to fend off appliance vendors like Netezza. Charlie Garry, who was at the time an independent analyst, told me Howard was all wet and that IBM is just playing catch-up with the likes of Teradata, which has been doing a similar thing for years and has a tight hold on the data warehousing market because of it.
All that’s besides the point. The real question for Oracle Watch is, of course, what's Oracle doing to fend off Netezza?
Oracle's people didn't respond when I asked them, but Netezza, naturally, was happy to give me a few thoughts. Vice President of Marketing Ellen Rubin said she thinks that Oracle just "hasn't figured out what the game is going to be for this stuff."
"One of the things we see, because we deal often with very large databases and we deal with companies doing more sophisticated analyses and doing iterations over and over, we find it's hard for Oracle to play," she said. "Their system comes out so much from the transactional model. A lot of things they've done is tune and optimize to try to get to some basic level of handling."
Phil Francisco, director of Product Marketing at Netezza, said the problem is that Oracle's in control of the RDBMS software, but not so much integrated with the hardware and storage vendors that would be part of a solution in an all-in-one software/hardware/storage approach to data warehousing—i.e., an appliance. "You have to look at the appliance as a holistic combination of all three elements," Francisco said.
Netezza's going up against Oracle Database 10g and to a lesser extent installed 9i deployments in sales calls. Rubin said companies have come to Netezza complaining of having issues in terms of Oracle's performance and scalability in a warehousing situation. "They're finding they've had to limit a lot of business users from running the type of queries they'd like to have access to," she said.
Rubin said that with analysis-intensive work such as fraud trending, where you have to join data sets from different places, once you get over a terabyte or two, Oracle customers find it's tough to scale. They can limit the number of people on the system or scale back in other ways, but it makes it hard to do things like trending.
Of course, Oracle data warehousing takes a lot of trained Oracle DBAs to eke the performance out of large Oracle implementations, as well. Then, once the setup is tuned to whatever's running today, something goes and changes, whether it's the organization acquiring another business or a move to another market. "Often that will happen, and the customer will continue to use Oracle for transactional needs, but they shift away for data warehousing needs and use Netezza instead," Rubin said.
I talked to Purandar Das, senior vice president of Technology Services for Epsilon. His company is running on Oracle versions 8, 9, 9i and 10g, with a total of between 35 and 40 implemented databases as client solutions. Das is still an Oracle advocate—he said that in specific situations, Oracle's openness is attractive, such as in small situations where Epsilon supports multiple clients on a single platform. Also, he said concurrency is better with Oracle, which works out in a large user population.
But he opted for Netezza when Epsilon was looking for a platform that would support some of the largest databases and that required extremely complex segmentation and analysis with very fast cycle times.
"In the past, because of the price associated with trying to build a traditional relational repository to support one of these, we've looked at alternative ways to build it," Das said. "Netezza showed us a simple and inexpensive way to do it." Savings are 40 to 50 percent over Oracle, he said.
OK, so there's some input from one of Oracle's data warehousing competitors and one of that competitor's customers. Oracle, let's hear from you: What are you planning to do to fend off the threat of these appliance vendors? Anything?
StreamBase Systems, a real-time stream processing vendor, came out today with new benchmarks, along with a new strategy and road map. The benchmarks clocked the software at 500,000 real-time messages per second on a single AMD Opteron chip, but since, realistically, nobody gives a hoot about benchmarks besides vendors themselves, let's get down to the nitty-gritty.
Namely, is StreamBase bumping into Oracle with its TimesTen product in sales calls?
Bill Hobbib, vice president of marketing, says no, no bumping at all. "With Oracle/TimesTen, people are taking data and they just want to store it someplace in memory for the purpose of querying it," he said. "StreamBase, we're giving a big application development environment to build high-value applications. It's a different space. TimesTen is 'Hey, I have a lot of data. I want to let people query that, in memory.' Whereas people are doing complex events processing with StreamBase. With TimesTen they're doing something very simple, just querying data."
I tend to expect big things out of StreamBase, given that it was founded by database legend Mike Stonebraker, father of the popular relational databases Ingres and Postgres. Now StreamBase's chief technology officer, he was also the founder of Illustra Information Technologies, acquired by Informix, which in turn was acquired by IBM.
Whether or not it's in the same space as Oracle's TimesTen, it will be fun to watch Stonebraker's new venture. The company today, June 20, launched a strategic initiative to evangelize the value of stream processing, and starting on June 19 it had a downloadable version of its software available for developers.
The launch of the new road map took place at the Securities Industry Association Technologies Conference in New York, and it's kind of fun: Dubbed "The Quest for The Da Vinci Coder," it gives participants the opportunity to download the StreamBase Developer Edition by taking a test drive through "a fun, interactive, and intellectually stimulating competition."
The video you'll find there is stimulating, but I'm not sure it's your intellect that will be stimulated—that depends on your gender and/or sexual orientation. I enjoyed the monk self-flagellating himself with a mouse, personally, and the promise that you'll be able to write code that "doesn't suck." At any rate, you can enter the contest to win thousands of dollars. Which doesn't sound as tasty as the $$$$$$million-dollar$$$$$$$ Ingres challenge, but hey, it's not like you have time to do anything besides sit around and write code for a startup, is it?
The fur is flying once again within Oracle's bug-eyed sales division, with the second Oracle Direct vice president out within 6 months.
This time it's Sue Hayden who's flying the coop. (Or getting tossed off the top of the barn, as the case may be.)
I viewed an extremely terse internal memo dated June 5 from Oracle Direct's Mark Aboud that said Hayden has decided her career would be better served at Monster.com, effective June 15.
"Sue Hayden has decided to pursue a career opportunity at Monster.com," Aboud wrote. "Sue has made a significant contribution to Oracle Direct over the past 5 years in expanding the team, evolving the model and delivering growth."
Wishes her all the best in world, blah blah blah, etc. etc. etc., and then Aboud says Hayden's peeps will report to him until Oracle Direct decides how to structure the Commercial Tech organization for FY07.
Sounds like this is direct fallout from a recent to-do about compensation. An earlier memo, sent May 30 by Amy Senew-Brown (in Aboud's office, as I understand it), told Oracle Direct reps that management had become aware of their rascally tendency to divide up larger contracts into multiple, smaller contracts in order to score higher margins on the sales.
The e-mail warns staff that this "unacceptable behavior" won't be tolerated. "Swift action, including termination, will be taken against any reps found to be splitting deals across two contracts for personal compensation reasons," it says.
Here's the text of the memo in full:
"This note is being sent as a general warning.
"Please be aware that is has come to management's attention that a handful of reps maybe splitting deals to ensure a large deal falls below cutoffs and are paid with a higher rate. This unacceptable behavior has gotten OD some negative press, used contracts precious cycles for selfish motives, and violated policy. Offenders will not be tolerated. Swift action, including termination, will be taken against any reps found to be splitting deals across two contracts for personal compensation reasons."
Selfish? Selfish? Oracle sales reps, selfish? Perish the thought! Negative publicity for Oracle sales? Gosh, that would be so very out of the norm, I can't bear the thought of it!
Snarkily delighted sarcasm aside, can we imagine that Hayden did something requiring "swift action" on the part of the database empire? Did she post the ol' resume on Monster.com and fortuitously stumble upon a port in the storm while she was at it? Dunno, but I'm open to fiddling with speculation.
The earlier VP departure came back in December 2005, FWIW, when Hilarie Koplow-McAdams, senior vice president in charge of Oracle Direct sales operations, quit to head to a senior job at Intuit. This came after an 18-year stint at Oracle, thank you very much, and after a career notable for her having headed up the direct sales and telesales effort.
Oracle's results for F4Q06 are coming up in the last week of June. UBS put out a note yesterday pondering whether the planets are starting to align for our buds in Redwood Shores, given that database license revenue shortfall could be made up by stronger than expected sales in applications.
I'm wondering about the database revenue figure myself. Oracle was at the top of the database dog pile in both Gartner and IDC's most recent numbers—no surprise there.
But as Gartner's Colleen Graham told me at the time, Oracle's both benefiting from and being punished by the push to put databases on Linux.
As Gartner reported, Linux grew the fastest of all the RDBMS platforms, at 84 percent. That growth was driven primarily by Oracle, as well as the maturation and acceptance of Linux as a mission-critical database platform. Oracle is the biggest Unix vendor, with 70 percent of the Unix market. Thus, RDBMS revenue on Unix may dip only 1 percent, but Oracle's license revenue from Unix dipped disproportionately at 3 percent.
Other things to expect: UBS thinks Oracle has spent the last four or five quarters rebuilding pipeline for the PeopleSoft component of Oracle's applications business and thus that double-digit organic growth is going to continue into 2007. This isn't surprising, given Chuck Phillips' recent reiteration at the OAUG conference in April that Oracle is sticking by its guns when it comes to upgrading its own E-Business applications, as well as PeopleSoft, JDE and Siebel apps, beyond 2013--the cut-off date it initially gave when it announced the Fusion road map in 2005. As eWEEK's Renee Ferguson reported, the announcement, according to Phillips, meant no forced upgrades for customers.
Oracle may indeed set customers' neck hair on end with its strong-arm sales tactics, but it's doing good by all the customers it has swallowed in the past few years. Smart move, and Oracle deserves to reap some bounty off the virtue of sticking by its word. And it will, the way UBS sees it: UBS expects Oracle stock to continue to outperform. From the UBS note:
"Our checks throughout the quarter uncovered continued momentum in Oracle's applications segment, both in the U.S. and Europe, as it is clear that the PeopleSoft and JD Edwards sales pipeline has finally started to rebuild over the past two quarters. In fact, our checks support our view that the divergent license growth that existed between SAP and Oracle over the past 1.5-2 years will continue to narrow over the next year on an organic basis (i.e., application license sales excluding Siebel as this asset has not been owned for 1 year).
"Our checks on the applications segment were stronger overall than what we uncovered for the database side of the house, where the quarter started off strongly as several large deals pushed from F3Q06. While our checks uncovered slow business signings in this segment, in particular in Europe in the middle of the quarter, over the month of May demand seems to have rebounded.
"In fact, we heard of increasing large deal activity over the final few weeks of the quarter, although we have yet to uncover whether these have closed. With the stock trading at 14x our CY07 EPS estimate of $0.99 (excluding ESO expense) we continue to view Oracle as a top idea in our sector for 2006. With continued improvement in applications execution, the nearing of easy database comparables (in August), the likelihood for further aggressive share buybacks, and the fact that we believe in the near term there are no large deals on the horizon (defined as deals greater than $1b), we view the stock as likely to continue to outperform this year as the multiple expands."
Oracle's losing Joseph Grundfest from its Board of Directors upon the close of Oracle's fiscal year on June 2.
Grundfest, who's been on the faculty at Stanford Law School since 1990, is resigning from Oracle's board to take on the job of co-director of the newly launched Arthur and Toni Rembe Rock Center for Corporate Governance at the law school.
Oracle's losing a lot of brains, here: Grundfest was a commissioner of the SEC from 1985 until 1990, and he's also the William A. Franke Professor of Law and Business at Stanford.
On Oracle's board, he was the chairman of the Compensation Committee and a member of the Finance & Audit Committee.
Beyond that, I don't know much about Grundfest, aside from the fact that Oracle will be losing the guy who said, in essence, that heck no, this ain't our last bid to buy our peeps at PSoft, during the Oracle/PeopleSoft/DOJ epic.
But Googling him does turn up a charming, typo-ridden, foaming-at-the-mouth, run-on masterpiece of a diatribe about Kuala Lumpur, money laundering, pandering to the Christian right and gagging legitimate business journalists from Twin Cities indy media. Here's an excerpt. Any reader who can translate this into English gets inducted into the Oracle Watch Hall of Fame:
“…And James Dale Davidson a founder of NAANSS and Genemax that was an illegal pump and dump scam also made the same fraudulent claim about Endovasc and Genemax being 'naked shorted' through his Agora Inc.'s 'Vantage Point' mail and internet tout network and so the fraudulent 'naked shorting' claim even dates back to 2002 and has a Stanford connection by way of those who promoted and dumped millions of shares of Stanford connected Endovasc shares through a Charles Schwab account.Principally James Dale Davidson and also the Texas attorney John M. O'Quinn made the fraudulent claim for a number of pump and dump scams including the Endovasc scam that Stanford and Christopher Heeschen and John Cooke profited from as well and,I believe,knowing it to be a fraudulent claim of the Stanford University connected Endovasc pump and dump scam to divert attention from their illegal pump and dump scam that may include money laundering with unaudited Endovasc shares as far away as terrorist money laundering suspect Kuala Lumpur and Dubai,UAE. Why ?.
“...It was only recently that Stanford law professor and ex SEC Commissioner Joseph Grundfest advocated for SEC or government confiscating business journalists' records and communications with informants in the business or suppposed 'private sector'.I believe this is an expression of far right and priviledged individuals to gag professional business journalists while leaving 'free speech' for insiders to various penny stock or other 'securities' scams for manipulation or insider deals to tout and lie and use the internet to set up their scammy penny stock touting websites claiming a 'naked short selling' conspiracy some even using Christian or Christiainty on their sites to cater to potential Christian religious suckers.Criminal fraud should never be seen as valid free speech no matter how politically or economically or legally connected the individuals.”
Yeah, yeah! What they said!
Duck, here comes SAP's Sapphire user conference, and the battling gusts of hot air from Oracle and SAP may knock you off your feet this week.
Oracle's hot air was that more than 500 customers have selected Oracle over SAP over the past nine months.
That's nice. I picked one of the companies, at random, from the list Oracle provided in the press release. I didn't recognize many of the company names, so I thought I'd just poke around and see what kind of companies are snubbing SAP.
After all, I'd read Renee Ferguson's eWEEK story on how SAP is trying to coax and wheedle users off R/3 and onto MySAP, and what I gleaned is that the specter of R/3 being retired, given a bed pan and sent to the Old ERP Nursing Home could conceivably be scary enough to chase some SAP users into Oracle's arms.
Not huge brand-name recognition on the list, though. One of the few companies I recognized was Remy Corp., and that's only because I thought it was the company that made disposable razors.
It actually does human capital software, for full-time staff, consulting, M&As, executive search and the like. (Oh, yes -- Remington, that's right. They’re the ones that make disposable razors.)
What I found interesting was that PeopleSoft was shopping Remy around as a case study back in 2004. Oracle's inherited the case study and thrown it on its own site.
Well, sheesh, I mumbled. Are these 500 customers ones who in any way, shape or form actually considered SAP? An Oracle spokesperson wasn't sure if any of the customers actually migrated, even though Oracle's press release has Chuck Phillips hyping the OFF SAP migration program. One would be forgiven for thinking that the list of customers was somehow connected to the migration program, being contained in the same missive and being uttered with the same Oracle president breath and whatnot. But then, one mistakes human capital companies for shaving companies, so what the heck does one know?
Analyst Judith Hurwitz was kind enough to point me to an interesting BusinessWeek article that claims that SAP is actually the winner so far in this fight.
"It has capitalized on the customer uncertainty that Oracle created with its acquisitions," the article said. "SAP says it has landed 240 former Oracle and PeopleSoft accounts, including the likes of Amgen (AMGN), Waste Management (WMI) and Samsonite. When sports-equipment maker Easton-Bell Sports decided late last year to standardize on one software maker, SAP got the nod. 'Oracle has been growing by acquisition. It will take a while before it's integrated,' says Sharon Nelson, Easton-Bell's chief information officer. 'SAP has grown organically. That made it the safe choice.'"
Still, Hurwitz likes to say that old software never dies. "If it works, they don't want to go to the extent of changing it if they don't have to," she says. And that's true even if new architectures such as SOA (service-oriented architecture) would give businesses more flexibility.
So take anything said by either company with a mountain of salt. SAP has been rearchitecting its software for the last seven or eight years to make it a lot more flexible and modular, Oracle’s sniping about SAP's proprietary ABAP (Advanced Business Application Programming) language notwithstanding, and Hurwitz said MySAP is pretty darn modular as a result.
Of course, you're always going to have customers shifting back and forth between the two warring enterprise software players, given the large number of customers they have in common. It probably doesn't have that much to do with open standards or modularity so much as a given customer just getting fed up with a given vendor and deciding to throw in the towel. And, of course, it has to do with sales, ka-ching.
"That's what sales people do: They go and try and take customers and get them to move," Hurwitz said. "It's all in sales incentives, practices, coming up with an offering that appeals to the customer for various reasons. I don't think there's any magic to it."
For its part, SAP's hot air has to do with the fact that it's planning to declare at Sapphire that IBM's DB2 Viper database is the "preferred and recommended database" for midmarket customers to run SAP. IBM's Bernie Spang, director of data servers, told me that migrations off Oracle and onto IBM are up 60 percent since last year, but IBM's not divulging the actual number of migrations.
Oracle's 500 customers choosing it over SAP? It's a bit of puffery. What would be real news? As Ferguson said to me, Oracle claiming that 500 people scrapped SAP for, say, Fusion -- that would be the big news.
Don't hold your breath.
Oracle introduced two new security tools on April 25: Database Vault, which restricts data access rights of even powerful users such as database administrators, and Secure Backup encryption technology, which ensures that even tapes that get lost or misplaced won't be readable by the wrong set of eyes.
I asked some security experts for some feedback on the two products, and the general consensus is they’re good ideas, particularly when you put the two tools together.
Dick Mackey Jr., a principal at security firm SystemExperts, told me, in fact, that he thought they look like “great” ideas.
“Two of the problems associated with databases are protecting sensitive data from administrative access and ensuring that sensitive data is protected in backups,” he said.
“Having the both issues addressed in the database product should be a real boon to organizations struggling with regulatory and contractual compliance. I'm not familiar enough with the design to understand how it protects against workarounds/hacks, but it looks like a good direction.”
Mackey was also impressed with integration into Oracle Database 10g, as well, given that we’ll get the glory of encryption without the drag of performance degradation.
“The integration of the encryption into that product should help significantly with performance,” he said. “Performance has been a real problem for anyone attempting to do field-level encryption in databases and in messaging. Many of the organizations we work with opt for hardware-assisted stream encryption to deal with the speed issues. Unfortunately, the stream approach doesn't protect you from your administrators.”
But Database Vault should, given that it allows you to lock down access rights of just about anybody, including administrators.
If anybody is testing the products or planning to roll them into production, let me know whether they live up to their promise.
Also, keep an eye on Pete Finnigan’s blog. He’s been fiddling with his site and hasn’t had time to opine on the new tools, but he says he plans to.
”If it doesn’t kill you, it makes you stronger,” Marten Mickos said.
The CEO of MySQL, with whom I was chatting earlier today at this year’s MySQL Users Conference in Santa Clara, Calif., was talking about Oracle’s buy of Innobase and its InnoDB storage engine back in October.
The snag of Innobase was a shocker. The transactional storage engine was and is at the heart of MySQL’s groundbreaking 5.0 release of its open-source database—you know, the version that brought the stored procedures, triggers and views that enterprises insist on having in order to take the database seriously. Yeah, that one, the one that everybody said would catapult MySQL into the enterprise.
Well, it didn’t kill MySQL, and this year’s users conference shows that it has indeed made the company a heck of a lot stronger. Why? Because Oracle’s buy, as it turned out, caused every database company on the planet to give MySQL a buzz, all of them offering to build MySQL a storage engine to replace InnoDB.
Six months later, after yet another Oracle purchase of a crucial MySQL storage engine (Sleepycat’s Berkeley DB), the user conference finds MySQL busting at the seams with storage engines. Solid was the first out of the gate, adopting its OLTP (online transaction processing) storage engine to work with MySQL Server. The prototype, which isn’t open source just yet, is called SolidDB Storage Engine for MySQL. It will ultimately be available under the GNU GPL (General Public License).
There were a mess of other storage engines announced, but the important thing is that MySQL is doing just swell.
It’s interesting, though, to hear, after the fact, how the Innobase deal went down.
As Marten tells it, Oracle President Chuck Phillips called on Oct. 7 and said, “Marten, I want you to know we acquired Innobase. This was an opportunistic move. We don’t know yet what to do with it.”
Huh.
We don’t know yet what to do with it.
See, I, and I’d say half of the Oracle user and observer universe, go around with the vision of Oracle as, well, a somewhat ruthless, bloodthirsty and voracious predator in the software industry. It’s just so easy to picture the company as Snidely Whiplash and a company such as, say, PeopleSoft as Nell Fenwick.
Imagining Oracle as a ditzy shopper makes the company seem, well, a lot more cuddly.
“We didn’t do it to slow you down,” Phillips told Mickos. “It’s business as usual. We want to renew the contract.”
Good for them, Oracle the Dudley Do-Right of the Canadian Royal Mounties’ Database Companies Division!
And did Mickos trust Oracle?
“Trust, but verify—it’s a Russian saying,” Marten told me.
So how do you verify that the world’s largest software company will do what the world’s largest software company says it will do?
“By sitting down and saying OK, let’s renew the contract,” Mickos said.
And that’s what they did. As we found out at LinuxWorld in Boston a few weeks back, the contract was renewed, unchanged. Hurray Oracle! Hurray MySQL!
But wait.
Did Oracle want to change the contract?
Did Oracle try to change the contract?
Marten stared at me.
Then he looked at his PR handler.
She stared back.
Tick. Tick. Tick.
The silence stretched out.
You will forgive me for thinking the answer was yes, Oracle did indeed want to change the license. In what ways, Marten is not saying.
Boy, would I have loved to be a fly on the wall during the contract negotiations.
IBM has unveiled its Venom data storage compression technology for DB2 Viper. IBM gave me a sneak peek at this impressive stuff a few weeks back, but I didn’t know the final, nicely poisonous name until now.
When I first wrote about the technology I posed it as a storage battle between IBM and Microsoft, but in actuality it's bigger than that. IBM says that Venom, which reduces storage hardware and usage by up to 70 percent, is really targeted at the 30,000 Oracle customers who use EMC storage and want a new, cheaper alternative to buying more EMC hardware.
Venom is a first in that it will enable DB2 Viper to deliver mainframe-like data storage compression to Linux, Unix and Windows environments. It works with row compression that augments Viper's value compression, index compression in multidimensional clusters and backup compression technology. IBM is claiming that it "eclipses" Oracle's older generation of table-based compression capabilities, yielding significant disk, I/O and memory savings, particularly for large tables with repetitive data patterns.
Beyond wanting to puncture Oracle, IBM isn't being coy about sinking its fangs into storage hardware competitor EMC.
EMC, with its acquisition of software companies such as Documentum, is increasingly competing with IBM in the software space, so this is a nicely complementary tit for tat.
DB2 Viper is due out in mid-2006. The open beta is available on IBM's site.
The following is not news, but you wouldn’t know it from the headlines flowing over the wire:
12:25 PM ET 4/18/06 [ORCL] ORACLE 'INCREASINGLY INTERESTED' IN OPEN SOURCE: PHILLIPS12:24 PM ET 4/18/06 [ORCL] ORACLE: DOESN'T BELIEVE IN 'RIDICULOUS PRICES' FOR ANY DEALS12:22 PM ET 4/18/06 [STT] STATE STREET SHARES UP 7.7% TO $64.8512:21 PM ET 4/18/06 [ORCL] ORACLE: WILL REVIEW 'SELECTED OPPORTUNITIES' IN OPEN SOURCE
During the question and answer session following Oracle’s conference today, during which it announced a buy of telecom vendor Net4Call and laid out its plan to basically own the telecommunications industry lock, stock and busy signal, Oracle CEO Chuck Phillips and Senior Vice President Tom Kurian were asked how open source fits into Oracle’s overall strategy in the middleware space.
Phillips answered by pretty much reiterating exactly what Larry Ellison said in the Financial Times interview from Monday. That’s the one where Ellison said Oracle has been pondering acquiring an open-source vendor such as Red Hat or Novell in order to control its own stack, operating system and all, except that he doesn’t want to go broke doing it.
“We don’t believe in paying ridiculous prices if, at the end of the day, you don’t own the intellectual property” behind such an acquisition as Red Hat or Novell, Phillips said in his most “I’m repeating what Larry said” tone. “We view open source as a friend of ours: something that in the past has been helpful to us in bringing us new users and in training new users we might not have otherwise reached and who, over time, become Oracle customers.”
As I said yesterday, one analyst told eWEEK that Red Hat and Novell are, indeed, too pricey for what Oracle would get, but that Ubuntu or Mandriva would make for smart acquisition targets.
My colleague Jason Brooks then blogged that there’s nothing to buy when it comes to Ubuntu, but that Oracle would be smart to start its own project based around Debian, similar to what Mark Shuttleworth did with Ubuntu: in other words, set up a foundation dedicated to putting a “more human face” on Debian, Brooks wrote.
“Ubuntu exploded in popularity because the Debian foundation on which it was built was already so strong,” he wrote. “Debian is already a major force in Linuxland, and a child distro of Oracle and Debian would pay back real dividends to the Debian project upstream, both in greater recognition of Debian's enterprise-worthiness and in actual code contributions from Oracle's distro team.”
Regardless of whether Chuck and Larry are hunched over their computers reading eWEEK’s sage advice, Phillips said we can expect … well, something … in the open-source space as Oracle’s strategy evolves. “We’ll have more to say,” he said. “But we’re increasingly interested in open source as a way of reaching more customers we can’t [reach] today.”
So. Ellison drops one of his Ellison bombshells on the Financial Times, along the lines of, Yeah, we want our own stack. Yeah, I thought of buying Novell. Nah, Red Hat’s too pricey.
So what happens? Red Hat’s stock suddenly becomes persona non grata, or monetis not worthitus, as the case may be. What a dis! Goldman Sachs lowered its rating of Red Hat stock to Underperform, based on the fact that Red Hat had the gall to purchase JBoss and thus start its crawl up the infrastructure stack, thereby leading it into direct competition with Oracle and IBM—two important partners of Red Hat.
“Oracle, in particular, seems likely to align with a competing Linux distribution in an effort to deliver a bundled, integrated, open-source infrastructure stack of its own,” the Goldman note noted.
“Red Hat Linux is already well entrenched as the dominant Linux distribution in important geographies such as North America, but a divided market could invite greater competition for Red Hat and more aggressive pricing," Goldman continued. “The JBoss acquisition is a catalyst event, signaling a fundamental market repositioning for Red Hat and recent comments from Oracle confirm our concern that Oracle now views Red Hat as a competitor. We suggest investors be more cautious on the stock near-term as Oracle unveils its own open-source stack initiative and are lowering our rating to Underperform.”
My esteemed colleague, Linux-Watch Editor Steven J. Vaughan-Nichols, is encouraging Larry to lay off the Linux distributions, to build his own darn Oracle Linux and to keep his proprietary-model mitts off of open source. His reasoning is that a) the mindsets required by the two business models are too different and b) if Oracle bought Novell or Red Hat, all its engineers would jump ship and you’d be left with precious few tangible assets.
“I'm not sure it makes any sense for Oracle to buy a major, or for that matter a minor, Linux distributor. Oracle likes using open source and Linux both within the company and as an operating system for its databases and applications. Ellison, though, is none too sure about the wisdom of paying big bucks for companies that don't own their product's intellectual property,” Vaughan-Nichols wrote.
I have to disagree on both counts, however. First off, as independent analyst and eWEEK contributing columnist Charlie Garry says, open-source companies aren’t about Birkenstocks and free love. They’re about making money.
“Let us not forget that MySQL, JBoss, Sleepycat and others were created to make money,” Garry wrote. “Even developers of open-source software have to eat, buy clothes and watch cable television. That all requires money.”
Let us also not forget the lessons we’ve been learning about Oracle’s dalliances in open-source companies. When Oracle bought Innobase and Sleepycat, I admittedly was one of the conspiracy theorists who thought Oracle was out to squash competition or something of that evil ilk. But when I talked to MySQL Marten Mickos at LinuxWorld a few weeks back, he told me that Oracle had renewed the license for the Innobase InnoDB storage engine, unchanged from preacquisition, and that MySQL is poised to release a new transaction storage engine of its own making at its users conference later this month.
I agree with Garry at this point: What Oracle wants is to make money with open source, not to crush competition. Well, OK, maybe to crush Microsoft competition, but really, MySQL has never been a competitive threat to the database Titan anyway.
“Without MySQL, what would Oracle do with Innobase? This has to be about growing the space and looking for new opportunities. Having another software heavyweight involved will only grow the open-source market, not hurt it,” Garry wrote.
I agree with Vaughan-Nichols: If Red Hat engineers fled the company after an Oracle purchase, Red Hat’s value would be negligible.
I don’t agree that Oracle has serious plans to waste its money on Red Hat or Novell, though. I’m more persuaded by the intriguing thought of Oracle buying Ubuntu: a hugely popular Linux distribution that could probably be purchased for a song.
Come on, lay it on me: I’m waiting to hear why that wouldn’t be the smartest thing Oracle could do and the real reason that Red Hat’s stock price was dismally downgraded after one happy week of flying high on the JBoss news.
