I work for Amazon Web Services. The opinions I share in this blog are my own. I'm *not* communicating as a spokesperson for Amazon.
In other words, I work at Amazon, but this is my own opinion.
Join 747 other subscribers
All content is © Kevin Closson and "Kevin Closson's Blog: Platforms, Databases, and Storage", 2006-2015. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Kevin Closson and Kevin Closson's Blog: Platforms, Databases, and Storage with appropriate and specific direction to the original content.
Kevin Closson's Blog: Platforms, Databases and Storage 
Hi Kevin,
Oracle’s announcement is here:
http://www.oracle.com/us/corporate/press/1563775
It seems it’s not certified for RHEL6 yet!
“Oracle Database 11g R2 and Oracle Fusion Middleware 11g R1 will be available on Red Hat Enterprise Linux 6 (RHEL6) and Oracle Linux 6 with the Red Hat Compatible Kernel in 90 days.”
Oracle says Oracle Database 11g R2 is certified for Oracle Linux 6 with the Unbreakable Enterprise Kernel.
Only three points I can make on that:
1) That’s what I get for speed reading
2) Oracle delaying RHEL6 until the end of their FYQ4 (90 days) is no surprise
3) I’ll be updating this blog entry June 1st and I can pretty much predict what I’ll be writing.
Hi Kevin,
Wim wrote some clarifications here:
https://blogs.oracle.com/wim/entry/lots_of_goodies
Oracle did not certify RHEL6, but it will be eventually certified when (if?) Oracle decide to certify the RH-compatible kernel…
The big news is you can use the security/erreta update service for free, and pay support only if/when you want support. So Oracle aims to replace Centos/Scientific Linux as the default free production grade linux….
Whether that makes business sense or just aims at hurting RH remains open…
Yeah, I knew about the 90 days. They have said they’ll do it. We’ll see the sincerity in about 89 days from now.
As for the move to provide free update service, I think that is a play to seed the market. I think it looks like fair play, but who knows how compelling it will be. In fact, let me say more on the matter.
For Oracle to make any move that is intended for nothing more than to “hurt Red Hat” is also fair play as long as the move is not rooted in anti-competitive practices. For example, Oracle could make OEL exceedingly popular in their accounts by giving free RDBMS licenses so long as the customer migrates from Red Hat on Dell servers to Oracle Linux on Sun x64 servers attached to Sun Unified Storage (a.k.a., ZFS Storage Appliance). That would be anti-competitive because it would reflecting using a monopoly position (RDBMS) to gain traction in a space they (Oracle) are clear losers (x64 and NFS filers). Don’t get me wrong, I’m not saying I have knowledge of Oracle doing so.
Since I’m on matters hypothetical, consider how anti-competitive it would be for Oracle (hypothetically) to sweeten a huge RDBMS licensing deal by, hypothetically, throwing in a couple of full-rack Exadata Database Machine configurations for free. While hypothetical, that would be clear abuse of a monopoly position in RDBMS software. Of course any such hypothetical customer would be crazy not to take the deal, moreover, even crazier to not use those hypothetical Exadata boxes for any up-coming hypothetical hardware refresh (e.g., Power6 to Power7 10g to 11g is in the plans). But, again, this is only a hypothetical example of a anti-competitive practice. I’m not saying I know first hand of such Oracle practices, nor am I accusing a company, with leadership of likes of Msrs. Ellison and Hurd, of promoting a business culture that would even consider any anti-competitive practice of any variety. In fact, I challenge anyone to provide a concrete example of any illegal business practices conducted by Oracle in the present or past because I’m not blogging the matter.
Kevin-
While we’re are all entitled to our opinions, Oracle certainly does not hold a monopoly position in RDBMS. By definition, they would have to be the sole provider, which they are not. In fact, Oracle holds less than 50% of the RDBMS market (by revenue) per the Gartner 2010 Worldwide RDBMS Market Share Report[1].
In addition, offering goods and/or services at discounted rates is quite common in the enterprise world, and nearly every vendor I know does this. This also includes your employer, EMC, the #1 worldwide storage software market leader [2] as well as the #1 leading provider of external disk storage systems [3]. Undoubtably EMC has tried to use their market presence to expand into other markets by way of dealing as well. There is obvious room for interpretation, but I see no clear anticompetitive practices by either Oracle nor EMC based on either the OECD definition [4] or Wikipedia [5].
Regards.
[1] http://www.oracle.com/us/products/database/number-one-database-069037.html
[2] http://finance.yahoo.com/news/emc-achieves-decade-storage-software-205700968.html
[3] http://finance.yahoo.com/news/emc-leads-external-disk-storage-212700745.html
[4] http://stats.oecd.org/glossary/detail.asp?ID=3145
[5] http://en.wikipedia.org/wiki/Anti-competitive_practices
Anderson,
You’ve come well-prepared with your comment and I appreciate that. Please note that my hypothetical wasn’t about using tactical discounts to effect a strategic move into a market. The hypothetical I gave was a dumping of the no-market-share product which, in this case, is Oracle’s x64 and NFS filer products. If the hypothetical (as I layed out) were ever a reality would that be a problem in your eyes?
Kevin-
It appears you are suggesting that the market share or popularity of Oracle’s x64 and/or NFS offerings somehow changes the properties of said actions.
If IBM were to hypothetically give away free DB2 licenses to a current Oracle customer if they were to purchase new IBM POWER servers to replace their current Sun hardware would it be any different? I’m sure such actions have never taken place, you think? Would you be equally as critical of IBM for such “anti-competitive” actions? Or would you not consider this to be the same?
This discussion seems to me to be more about showing animosity toward Oracle and its products than an unbiased discussion of hypothetical sales practices.
To answer your final question, no, I do not see it as a problem. If any company wants to offer me discounts for solutions that meet my organizations’ needs, as you said, I “would be crazy not to take the deal”. I say that despite how popular the current solutions I have from that vendor are and how unpopular other solutions may be — popularity/market share is irrelevant, in my opinion, if it works.
Anderson,
These words don’t reflect animosity towards Oracle’s products. The word “if” is at the center of the discussion as it tracks back to Ofir’s installment in this comment thread. The only animosity I hold towards Oracle is directed it Oracle’s business practices, not products. That is my opinion and I feel entitled to hold such opinion. But, again, this is a discussion of the hypothetical.
The problem with the comparison to an IBM pot-sweetening hypothetical is that Oracle’s RDBMS revenue position is just shy of 50% of the market and more than all the 5 closest competitors combined. If IBM were to throw DB2 in for free for a migration from Oracle+SPARC to DB2+Power I would not consider that anti-competitive. Given the position DB2 holds in the market (one of the “clump of five”) I’d see it as nothing more than a sort of discount offered to someone purchasing Power gear. It is in no way the same as the hypothetical Oracle scenario I drew. For review, there is no parity between the two hypotheticals in my mind:
1. Existing Oracle RDBMS customer gets free Exadata if they renew a huge RDBMS licensing contract
2. Existing Oracle+SPARC customer migrates to DB2+Power where DB2 is thrown in for free
The reason I feel these are not the same is in #1 the dominant product remains dominant and the freebie is using the leverage of that dominance to muscle into a market where there is no appreciate presence (Sun x64 gear+Storage). On the other hand IBM would be throwing in a market lagging product to sweeten a Power sale. I see that as no more than a discount. Now if it were IBM throwing the Power server in for free and collecting a little revenue on the DB2 licenses that would be a different story because Power (last I checked) has a dominant position in Unix server revenue. Now having said all that, the Unix server market is a rapidly dying market whereas Oracle RDBMS is dominant in a growth segment.
Well, it’s likely, Anderson, that you and I are the only ones reading this so the philosophical exchange is probably moot.
P.S. I stand corrected on my use of the word monopoly in my original hypothetical as a result of your previous installment in this comment thread. I was wrong. Oracle’s vast dominance in the RDBMS segment does not constitute a monopoly.
P.P.S. I still do not think Oracle making a fresh offering of free errata update service for OEL constitutes any sort of malpractice whether or not the move is indented to put pressure on Red Hat’s dominance in the segment. It is a move to add value by an eager underdog player that has no appreciable presence in a segment of the market. It is fair play and could prove beneficial for customers.
Anderson & Kevin:
No you guys are not the only ones reading this thread. I have followed it with interest, and feel I must now wade in.
Anderson, I am not sure what your background is, and so cannot comment on your ability to define accurately a term like “monopoly” as that term is used within US antitrust law. However, I am an attorney, and practiced antitrust law for many years, before turning to IT, where I now as work as an Oracle-focused storage professional. Like Kevin, I work for EMC.
The situation with Oracle as it presently stands is very similar to the situation that existed with Microsoft in 1998 when the US Justice Department initiated legal action against Microsoft. In that case, it certainly was not the case that Microsoft had a 100% market share in either marginal sales or the installed base of computer operating system software. Certainly, the Macintosh was a viable alternative, and UNIX and Linux were both coming on strong. Also, there was lots of legacy minicomputer manufacturers still in existence then.
However, the US Justice Department determined (correctly in my view) that Microsoft was exerting excessive market power by virtue of their overwhelming competitive advantage in the area of operating system software.
In many ways, the situation with Oracle is even worse: Operating system software on desktop computers is certainly an interesting and critical market. However, RDBMS software is the fundamental underpinning for our entire economy. Virtually every enterprise of any size runs its entire business on RDBMS software, making it the single most business critical and strategic category of software technology in the world.
Further, the surveys of folks like Gartner dramatically underestimate Oracle’s market dominance in the area of RDBMS software. This is because Garter surveys count license sales. Oracle is shown by Gartner at around 50% of the market, according to the latest stats I have seen. IBM is less than half that at around 22% with Microsoft in third place at 17%.
The problem with these statistics are twofold:
1. The IBM market share is largely legacy. Current new adoptions on IBM DB2 are exceedingly rare. Thus, although IBM has a significant presence in the market, it is in a relatively small number of sites with very large databases, most of which is on legacy platforms.
2. In contrast, Microsoft has a very inflated market share, due to the nature of the typical Microsoft SQL Server configuration, which is dramatically smaller and less mission critical than the typical Oracle configuration. Thus, in terms of market share impact, the effect of Microsoft SQL Server is overstated.
A much better way to look at the relative market power is in terms of market cap. Microsoft is the largest single software company in the word with a market cap of around $260 billion at the end of 2011. However, over 90% of that market cap is attributable to two products: Windows and Office. That means that all of the remaining products of Microsoft (including Exchange, SharePoint, SQL Server, etc.), only amount to around $26 billion in market cap.
Oracle’s market cap at the end of 2011 was $144 billion, almost all of which was attributable to its server software products. (Most analysts believed that the Sun acquisition had a negative impact on Oracle’s market cap at that time.)
Conservatively, in terms of market cap, Oracle RDBMS software is worth 10x of either its competitors. That is the level of market dominance we are talking about.
The tactics of Oracle in terms of, for example, offering a $1 million discount on RDBMS software in exchange for a $400K purchase of ExaData software is a practice referred to in the antitrust law as dumping. Oracle is using its market dominance and monopoly position in the area of RDBMS software to increase its market share in a market in which it is competitive: In this case storage hardware. This is completely comparable to the basis for the action against Microsoft (dumping Internet Explorer in competition to Netscape) which resulted in the action in 1998 against Microsoft.
I believe that the time is ripe for a legal action against Oracle on these grounds, and I fully expect that to happen relatively soon.
Monopoly was correctly used, in the contextual sense of monopolistic practices (Anderson’s definition is the economic definition, not the legal definition). That’s why in April, 2009 you have IBM, Oracle, Nokia and Sun joining the European Commision, Google and Mozilla in accusing Microsoft of monopolistic practices by bundling IE with Windows… then in September, 2009, the EC investigates the coming Oracle and Sun merger… in 2010, another problem was avoided by making java an oligopoly… and it goes on and on.
Market cap is very bad indicator for this purpose, since it is so intertwined with ROI assumptions (for example, interest rates and competing investments). Absolutely nothing can change in reality, and capitilization can vanish with those parameters. That’s what FUD is all about. People avoid Pink Slime, never thinking regular old marshmallows are much more disgusting. It is a good way to show the relative value of various products within vendors, showing a problem with Gartner methodology, though, the problem is in going further.
As far as new licenses, I think it is a good point, but I also think all have recognized the growth will be elsewhere, it’s only gravy, temporary on a mature market. Milk it while you can, but it’s only a tactical advantage, not that there’s anything wrong with that.
There will always be legal action, that’s just a cost of doing business. The trick is to keep it a cost, rather than a driver. IBM had a consent degree in 1956, did that stop it from becoming a behemoth? Did the breakup of AT&T in 1982 stop it from putting itself back together? Back in the ’90s, some journalist (I think Stewart Alsop) asked people to predict IBM’s market value in 2000. Apparently, two people said “zero,” and I know I was one of them. The best investment is a well-run business, and as much as we would like to have our heroes and hope everyone else fails, with very large companies there is a lot that can happen. Operating systems and database engines are really the most minor part of it. Even patents expire.