Slashdot Mirror
Oracle to Compete With Red Hat for Linux Support
PCM2 writes "It's not Oracle Linux, but Larry Ellison has announced that Oracle will be providing full enterprise support for Linux. This means not just phone calls but also patches, security fixes, and backports, in addition to indemnification from lawsuits like SCO's. This puts Oracle in direct competition with its erstwhile partner, Red Hat, whose entire business is based on providing similar support for its Linux distro and related software."
Perhaps just a strategy to get RHAT stock low enough to buy them?
Vote Libertarian
``Oracle will be providing full enterprise support for Linux.''
That should have said "Red Hat Enterprise Linux 3 and 4". They're not supporting every GNU/Linux distribution, nor are they supporting just the kernel.
Remember, there is no such thing as the Linux operating system. Linux is just the kernel, and the various distributions based on it are all different.
Please correct me if I got my facts wrong.
Why they would want their own distribution (either buy Rhat or make their own)..
But why not use established systems with guaranteed update mechanisms? Something like Debian with the stable branch comes to mind.
Ok, in summary, Oracle will support only RHEL3 and RHEL4 distributions (per TFA). They didn't mention CentOS and said they wouldn't be packing it themselves, so the implication it is only copies purchased from RH. Best I can figure is that Oracle would be offering the equivalent of an 'extended warranty', targetting those who took the shortest support contract possible from RedHat and paying oracle with the rest, meaning either their hoping their name will carry weight or they plan to undercut RedHat for long term contracts.
TFA says RedHat doesn't sell the 'OS', but that's bullocks. You cannot legally get RHEL without paying for it (some of the copyrighted artwork and name), hence the whole point of the existence of projects like CentOS. Their fundamental business is built on support, but it changes not the fact that they do not give away the distro they sell anymore.
XML is like violence. If it doesn't solve the problem, use more.
Because Oracle commits tons of code to the mainline linux kernel, unlike RedHat....
/usr/src/linux /usr/src/linux
Oracle seems to support Oracle - like ocfs2, which so far as I can tell, is the only substantial Oracle contribution in mainline.
grep -r oracle.com
grep -r redhat.com
RedHat has invested in major contributors by putting them on staff. Oracle? Not so much.
If it's anything like their support for their flagship products, Oracle and PeopleSoft don't bother. Anytime we report an issue with our multi-million dollar enterprise implementation, they spend several weeks trying to find some other party to blame. It's your hardware, no it's the network....etc.
I will not be pushed, filed, stamped, indexed, briefed, debriefed, or numbered. My life is my own.
Oracle's internal organization and practices lead to expensive services. They probably can't compete with Red Hat on price.
My guess is that Oracle isn't really targeting Red Hat, they're targeting IBM and eventually Microsoft. Larry E. isn't noted for humility and, if he takes out Red Hat, it's just a way station on the road to a greater goal.
Ellison explained it more fully in his presentation. Basically, Oracle will provide Enterprise support for RedHat as well as its own version of Linux called Unbreakable Linux. Unbreakable Linux will essentially be a copy of RedHat. Every time RedHat comes out with a new release, Oracle will take the source code, remove all of the RedHat trademarks, and release it as Unbreakable Linux. Oracle will allow both RedHat and Unbreakable installs to get support from them, including bug fixes, patches, and other software.
All a RedHat client has to do to move to Oracle support for their RedHat install is to stop using RedHat's update servers and start using Oracle's. Oracle will provide patches, and will backport those patches to earlier revisions than RedHat does in order to keep enterprise-level clients from having to upgrade all of their systems too often.
So, as it stands now, Oracle will basically be offering a higher level of support than RedHat for the same exact software for a lot less money. RedHat is going to be forced to drastically reduce their prices just to compete.
I would think that over the longterm Oracle's Unbreakable Linux will fork off, especially if this ends up seriously damaging RedHat, but for now Unbreakable Linux is nothing more than a re-branded copy of RedHat.
RMS always argued that free software is pro-capitalist, because there is a free market for support. I think it's great that we're seeing this argument validated with real-life examples.
Yes, there are several vendors who support their own distro of Linux, but are there previous instances where a third party (Oracle) is competing with a vendor who itself does support (RedHat)?
RedHat is stagnant. I have to admit a degree of ignorance here, but as far as I can tell, they are not really doing anything to excite interest in their market. They offer support and their own distribution (apart from Fedora Core) at outrageously high prices. (Even Windows server solutions are cheaper than RedHat.) Sure they gobbled up JBoss, but I do not think there is as much market overlap as one might suspect.
I might even go so far as to say RedHat has done a fair amount of damage to Linux adoption: they create high costs and little value or innovation likely because they face no direct competition.
With Oracle entering the picture, RedHat will be compelled to move quickly—to at least do something. I am not even quite sure what that is, but one way or another, this is adds choice for the market and that is always good, whether it results in a better RedHat or no RedHat.
Why bother.
``Where do you draw the line?''
If you're talking about a specific distribution (which is often the case), use the name of the distribution. If you're talking about the kernel, say "the kernel" or "the Linux kernel", and if you're talking about GNU/Linux systems in general, say "GNU/Linux", to avoid ambiguity.
``Linux distribution is a well accepted term and the practice of shortening it to Linux is well accepted.''
I agree, and I accept that usage, but, in this case, Oracle is supporting a single distro, not Linux in general. Saying that they support Linux is, at best, unclear, and I would say false.
``Without the GNU toolset (or one like it), the kernel would be essentially useless.''
I suppose you mean to suggest that we shouldn't be saying "GNU/Linux". However, I've built and seen systems based on Linux that didn't include GNU software, as well as systems that included GNU software, but not Linux. I can tell you that much of the identity of GNU/Linux comes from the GNU part, not the Linux part. That's why I prefer to use the combined term.
``If a desktop system, at least X, and generally Gnome or KDE is needed, so do you have to say Gnome/X/GNU/Linux in that case? If it's a particular config of a web server do you have to say Apache/Postgresql/PHP/GNU/Linux?''
I am not about to declare that everyone _has_ to call it a certain way, but I do like people to be clear, precise, and truthful. Konqueror is part of KDE, not Linux. Firefox runs on top of GTK, not necessarily Linux or even X. glibc is part of the GNU system, and works with various kernels besides Linux. Drivers for Linux won't work with AIX, no matter how many GNU utilities, X servers, and GNOME's you install.
In cases where it's relevant, it may make a lot of sense to describe a system as Apache/Postgresql/PHP/GNU/Linux, although the various components probably matter to different people. As a webmaster, I probably care about Postresql and PHP, and perhaps Apache, but not about GNU and Linux. As the sysadmin, I probably care about all of them.
Please correct me if I got my facts wrong.
I have not checked the prices of comparable RedHat and Microsoft server offerings. It turns out that RedHat is still cheaper, but by a trivial amount. Compare the RedHat Store (see: Server Operating System Products) and Windows Server 2003 R2 Pricing. (Wouldn't it be nice if Slashdot support post annotation or editing?)
At any rate, Windows might still be a superior server platform thanks to the effectiveness of ActiveDirectory, fine-grained ACL, and so on. I am no Windows apologist (on the contrary, quite the advocate of open source solutions), but I fear Microsoft may be leaping far ahead of their competition in this space.
Why bother.
I would think that over the longterm Oracle's Unbreakable Linux will fork off, especially if this ends up seriously damaging RedHat, but for now Unbreakable Linux is nothing more than a re-branded copy of RedHat.
Dystopian future: because Unbreakable Linux is built off RHEL (like CentOS is), Red Hat lose (some/half/all) of their support customers to Unbreakable, can't afford to keep producing RHEL, and Oracle base future versions of Unbreakable on what, now?
Sounds to me like they're trying to smoke Redhat out. Undercut their pricing until they have to drop their own pricing to the point they no longer make money. Then it's simply a matter of purchasing Redhat at the new low low discount price.
Oracle, unlike CentOS, has money. They hire the fleeing Red Hat people and eventually buy the shell of what was Red Hat for a song. Larry, being partial to bowlers, adopts a new logo.
Learning HOW to think is more important than learning WHAT to think.
what exactly does oracle get by buying red hat?
really. think about it.
i should say; what does oracle get by buying red hat that would still be worth anything after the purchase? after the employees flee their proprietary overlords and the red hat brand has been subsumed in the giant sucking sound of oracle's corporate engine?
the death or purchase (same thing, really) of red hat does not benefit oracle in any way. and it's going to be interesting to see if oracle can actually deliver linux support that anyone wants, even if it is cheaper. compare the customer satisfaction for the two companies and make up your own mind.
-- Moderation in all things, exceptions to all rules --
"The distro's just fantastic, that is really what I thinks, oh, by the way: Vista stinks..."
Get thee glass eyes, and, like a scurvy politician, seem to see things thou dost not.--King Lear
In this case, we have the following scenario.
Company A develops linux distribution, supports it.
Company B simply compies Company A's work, supports it as well.
My question is this, what is company A's incentive to develop a distribution? Because the development costs are 0 to company B and substantial to company A, company B can easily undercut the price of company A. It would seem like you'd have to be a fool to develop a distribution, since the GPL forces you to surrender your work to competitors who can easily undercut your price.
E = m c^3 Don't drink and derive E = m c^3
And of all of that is terrific, but none of it is supported by the vendor. You can call RH with your MySQL or PostGreSQL problems, as the primary thing they are selling you is support and updates. You can't call MS, you can scarcely get support form them on the features included in the OS without shelling out extra $$$.
If support and updates are unimportant to you and you're willing to run things like Cygwin and Apache on your Windows server to avoid paying for Red Hat, its a lot more likely you're just going to run a free as in beer linux distribution which provides the same tools with no support.
once you go slack, you never go back
> Oracle to Compete Wth Red Hat for Linux Support
Finally! Now you too can own a $30,000 version of Linux.
* In keeping with the Oracle tradition of no GUIs, no KDE and GNOME will not be included. Larry has his fingers crossed someone else will write one.
* In the Oracle tradition of installers written in Java, you too can have a relaxed day of installation watching those damned applet windows keep redrawing themselves... very slowly.
An interesting point from Dave:
And this quote made me laugh:
We aren't a full linux shop - we are a typical midmarket corp IT shop - running Oracle eBusiness Suite, Oracle DBMS, Oracle App Server, along with a variety of apps backed by either Oracle DBMS or SQLServer. Our infrastructure is a mix of RH linux and MS 2003 Server. All the Oracle products are on RH linux, and we have been paying Oracle and RH for support.
So, now Larry is telling me I can stop paying RH for support, and I can pay Oracle. My cost will be about 1/3 what I'm paying now to RH. When I call for support on one of my Oracle apps, I don't have to worry about whether it is a bug in the app, the DBMS or the OS - the support call is the same and they need to help me figure it out.
Where's the downside for me? If you aren't currently an Oracle customer - fine, keep paying RHAT for support. If you are an Oracle customer, it's a no-brainer.
What Oracle gets by buying redhat is experienced redhat linux *people*. In a service industry people matter. People talk to clients. People know what works and doesn't, and can ask the right questions. Really, Oracle is buying *knowledge*. If it helps, consider buying redhat as the purchase of interactive organic knowledge agents. Hehehe, well I laughed ;-)
Ah yep:
Unfakeable Linux
Don't you have someone you'd die for?
From: http://edelivery.oracle.com/linux/
It seems a rebranded version of RHEL (a la CentOs). -- Ernest
"The vendors aren't offering indemnification, Ellison said, and because of SCO, there's all this uncertainty and doubt about intellectual property. He says he will offer indemnification. In the Q&A at the end, he was asked if Oracle was planning to buy SCO to bring that uncertainty to an end. No, was the answer.
"Red Hat has a separate indemnification policy. In Red Hat's case, this policy is called the Open Source Assurance program."
Presumably if Larry really believed the SCO case had any validity he wouldn't even consider using RHEL. And in relation to RHEL and the GPL what's stopping anyone buying a single copy of Oracle Linux and repackage it and selling it with support contracts. Presumably if Larry doesn't allow this then Oracle is in breach of the license.
"We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software"
davecb5620@gmail.com
...Then this will be a shocking wake up call to Oracle. Oracle don't do 'communities' and 'sharing' so well. Most of the best documentation on their products is inside Metalink, hidden away from the prying eyes of the less than 'enterprise' customers. I'm an Oracle DBA, and I think that Oracle Database is a fantastic product, which is ultra reliable, scalable and well thought out. Support is generally pretty good, and it damn well ought to be for $25,000 a CPU! Contrast that with Oracle Application Server which is a buggy, badly supported, badly thought out mesh of acquired code and Oracle 'improvements'. Since it's written in Java, it's quite easy to pull OC4J apart and look at the code if you so desire. We run an enterprise Java application on Oracle application server, and it is a _lot_ of trouble. Try it on AIX, with the IBM JDK. Even more fun to be had there :)
The trouble is, Oracle don't know much about Linux really. They don't contribute much to it, and this move is really riding off the back of other people's work. It's more about taking back what Redhat took from them - a decent and popular application server product. Why were they so keen to get hold of Jboss? I think it's proberbly because their existing application server is so shocking...
Oracle do some things well, but they don't do community support, sharing, or cheap products. If they truly are looking to take over Redhat, then that will be a sorry day for Linux.
I think Oracle have bitten off more than they can chew. I'm sure you'll be able to get 24x7 enterprise support, but it will be offshore, and you'll have to speak to 6 people and an account manager, 6 more people, a customer services representative, fill out a survey and then finally produce a test case to prove that there is a bug in their product, which would be obvious if you could actually speak to a devloper who spent twenty seconds looking at the code. Even IBM do Linux better.
Where is the data to back Oracle's claim that they are better than Red Hat at supporting linux? The essence of Oracle's claim is that Red Hat has failed to provide true enterprise level support, and now Oracle is going to step in and do it right (and cheaper too).
But remember that CentOS already provides bug fixes and updates for free, so the part about being cheaper is basically irrelevant -- anybody who cares about cost can already get basic support for free with CentOS.
Also remember any good changes Oracle comes up with must be GPL'ed, so Red Hat can pick them up immediately and use them as appropriate. (this is a good thing in both directions!)
So we're really talking here about extended support -- complex problems where you have to pay somebody to care about your twisted problem and help you figure it out in detail.
My experience with both companies is that you have to pay big $$$$ to get anybody to personally care about you in any way -- even a hundred Red Hat licenses don't amount to squat in getting somebody personally involved to help solve your problem, and I have no idea what level of $$$$ you have to pay to get somebody from Oracle to care (at least metalink makes no claims to care about you).
So I rate both companies about equally poor in serving my mid-sized company when complex problems arise.
In terms of interactions we do have, Red Hat seems generally earnest and honest, while Oracle seems arrogant and greedy. We get shaken down by Oracle once per year -- basically the sales guy shows up once a year with a baseball bat aimed at your kneecaps, asks how much you can afford to pay, and then invents some ridiculously large number that we now have to pay.
So the background summary is nobody can do complex support cheap, and in general attitude I prefer Red Hat.
The particular question that comes up here, though, is why would anybody believe Oracle has any specific expertise to support linux at any complex advanced level? This is a hard thing, and Oracle has no demonstrated competence at it... who would bet an enterprise production server on that?
Red Hat has built up its reputation and competence over a dozen years, and devotes 100% of its energies to doing linux well. Oracle has an egomaniac leader whose company has developed exactly one good product (not counting recent acquisitions) and who claims without proof a world class expertise in supporting linux, a task he claims is trivial and therefore clearly does not understand.
Oracle has a long trail of failed initiatives behind it, and this smacks of another -- dabbling in a field where it claims expertise but truly has none.
What big company is going to place a critical bet based solely on Oracle's reputation in a field where Oracle has no expertise? Anybody sensible is going to wait a couple years and see if Oracle can pull this off.
The people trashing Red Hat's stock right now are forgetting that this is a hard business, and Oracle is just a dilettante.
There is nothing magic about Red Hat that makes them a market leader -- the company has simply put in the hard work over the years and developed the expertise and critical mass to be the leading commercial linux distribution and support company. Oracle is not going to out-do Red Hat on any of those points because it's not their core business and never will be.
I predict utter failure for Unbreakable Linux beyond the Oracle software stack -- after a couple years of dismal sales (hello Novell) Oracle will end up supporting linux internally as part of its own bundled software stack.
I can't think of a single reason why anybody outside the Oracle stack would go with Oracle linux over a combination of CentOS (price) and Red Hat (expertise).