Guest post by InfoWorld’s Savio Rodrigues views Red Hat’s notable marketing shift from low cost to high value as an “important shift in the ongoing evolution of open source software vendors’ business models”. Long left in the low-cost ghetto of enterprise IT mind share, RHEL is being pushed for its technical innovations, performance enhancements, and customer-requested improvements, rather than as a solution for cash-strapped shops. This shift, and the underlying improvements to RHEL 6.0, give Red Hat a legitimate shot against Microsoft, and open source a new model for competing with proprietary products. After all, focusing on low cost unnecessary limits the growth of open source business, creating a ‘low-cost ceiling’ that indirectly dissuades many IT shops from considering open source products.
Is about time Linux gets paid
Red Hat can talk up value all they want but Oracle can offer the exact same software and charge less for support.
Oracle has a long term strategy against the race to zero which is tie open source to hardware and their proprietary database.
The more you embrace open source the more you open yourself to being undermined by a large corp. The hybrid model is much safer for smaller companies.
They have been trying this for the past few years and its not exactly worked out.
The problem Oracle face is similar to the one that Ubuntu faces in the enterprise – when you have problems, would you want to go get help from the people who actually did the work, wrote the software etc or some middle company who will then wait for a fix to arrive from elsewhere (Which Oracle MUST do anyway, as it wants to have 100% compatibility with RHEL.)
As for mixed source being a blessing, it can cut both ways – it may benefit the bottom line of a community, but as the people relying on Mysql or SUN products to see if they are happy with the results of their practices.
It doesn’t have to work out in the short term, they have a safe profit generator called Oracle DB.
I wouldn’t call it a failure either given the customer list they have built up.
They’re playing a long term game and they have a much better plan than Red Hat.
Well Oracle is not 100 percent behind linux. they have proven that time and time again. + Please look at RHEL 6 and then point me to an equivalent OS release from Oracle (no, incomprehensible linux, or whatever they are calling there distro these days, is no competitor )
And what we are also starting to see is that Oracle hates itself and wishes to meet a quick end by going after all the stuff that the market enjoys and killing it (Android for example Solaris for another. Its also a known fact that Oracle hates christmas)
It doesn’t matter what RHEL adds which is the whole point. Any software they add to Linux can be taken and offered by Oracle.
Oracle doesn’t undercut Red Hat in anyway that CentOS doesn’t already do. If someone wants an enterprise ready linux distro there are plenty of ways to get it for free. However if your going to shell out the big bucks for a support contract its hard to argue against Red Hat since they are pretty much the best at that right now.
Oracle can make some arguments about DB integration of course or maybe through in some package deals or something. They are a big company with a lot of options, but they can’t just kill Red Hat that easy.
One could argue that Red Hat works with the likes of Dell and HP then with the consulting services and so forth provide the necessary support but with that being said the person whom you’re replying to does have a point. I don’t think that Red Hat will ‘die’ but I do think there will be a limit to their model some time in the future as HP and Dell take over more of the consultation side of the equation leaving Red Hat with software sales and support which is limiting in terms of growth.
In the long term I could see Red Hat being bought out by a big OEM; maybe Dell, HP or even Lenovo having an in house operating system that is hyper tuned and optimised for their hardware then work with middle ware vendors to tune and tweak their software to sit on top.
Personally for me I’ve never been a fan of the piecemeal approach of slicing the different parts up into different businesses because it ultimately leads to a tension between the competing parties and ultimately someone loses out in the equation to the detriment of all and the consumer.
I believe you don’t get Red Hat business either.
Red Hat does not sell software that much, it’s not Microsoft. It sells Service. It is competing with IBM, Andersen, Cap Gemini. They provide solutions on demand. The customers get exactly what they asked for. They don’t get a one-size-fit-all software and adapt their business to use it. They get the software that works for their business. The core business of Red Hat is the services that comes with the software, not the software itself.
There is no way Dell or HP can get software from Red Hat and provide the same level of service at the same time. If they want to provide the same level of service they have to fork it, and forking Red Hat costs a ton of money. They would end up being at least as expensive as Red Hat. But that is the price to pay. You can’t adapt the software and let it stay the same at the same time. Either you fork or you don’t provide service on demand.
Oracle has a list of customers that have switched and plans on taking more.
CENT doesn’t provide support, that is a major difference.
Undermined how exactly? And what would the benefits be with proprietary code in avoiding this?
Undermined by taking your innovative software and reselling it under a corporate label.
Oracle can’t take proprietary code and add it to their Linux stack.
What software would Oracle be ‘reselling? Afaik Red Hat doesn’t sell software, they write software and sell support for it. Despite your wishful thinking I just can’t see Oracle offering anything near the technical support Red Hat can offer given that they actually write alot of the code they sell support for.
Secondly I haven’t seen any tendences of Red Hat losing ground to anyone in the Linux Enterprise space.
Even through this recession Red Hat has done great with continued profits and revenue growth while most of the software industry has suffered hard.
It doesn’t matter if it is support that is being sold since the result is the same. Red Hat adds value and Oracle duplicates it.
How am I engaging in wishful thinking? Red Hat only writes a fraction of their own distro. Linux is a collaborative project. Oracle can support RHEL or any other distro. All Red Hat can rely on is the *perception* that it is better to pay a premium for the original source. How long can they depend on perception?
Compared to other Fortune 500 software companies their profits are paltry. They have a successful business but Oracle has a better long term plan which is to tie their own version of Linux to proprietary hardware and software so they can add value without it being assimilated by competitors.
You don’t get it at all. Let me give you an example. In my company, we have everything. We have CentOS, Oracle linux, Debian and RHEL servers. The CentOS and Debian servers are for non business-critical servers. The Oracle linux servers are for testing. For critical servers you can’t get around Red Hat. What if my server crashes because of a bug in whatever critical component? We loose several millions of dollars each day of downtime. So what we do? We call Red Hat, they provide a fix in the hours and we get our server back in hours. Depending on the contract, they can also pay for the downtime. OK, now let’s say I had Oracle linux on my critical server. Now let’s say it crashes. Call Oracle, they tell me that it is an upstream problem in RHEL. Can you fix it pretty please? No sorry but you have to ask Red Hat. Call Red Hat, they tell me that they can’t support because it’s Oracle linux. they will fix the problem when they have time sometime next month. Right, when the fix is available it won’t matter because all of my customers will have moved away.
Oracle linux does not replace RHEL. It’s not the same level of service. The support is level 0. They will help you install, provide training but they won’t help you fix the critical problems. It’s nice for some tasks but it’s not for critical servers.
So these organizations aren’t running critical servers on Oracle?
http://www.oracle.com/us/technologies/linux/025990.htm
Only a fraction of RHEL is actually written by Red Hat. It’s a Linux distro that is favored by enterprise for compatibility and its conservative development cycle.
I also don’t see why you seem to think they are incapable of fixing RHEL code:
Ellison also said, “We [Oracle] spend a lot of time finding and fixing bugs in Red Hat Linux, and we have no problem with that–we do that with lots of operating systems.”
http://blogs.computerworld.com/16997/oracle_rips_red_hat_and_sort_o…
It is open source after all.
If you don’t think Oracle support is good enough then that is fine but plenty of companies do. Oracle is positioning themselves as a one stop shop and they will be able to add value that Red Hat can’t duplicate.
If they were able to add value to their linux server, that would be fine with me. If they could fix bugs and implements features on demand, that would make them real competitors with Red Hat.
If they did that they would increase their cost them more and they would be just as expensive as Red Hat. You get what you pay for. I did call Oracle and I had to face their support team. Metalink is a bad joke when you compare it with the support you get from Red Hat. They just don’t have the same level of expertise and reactivity. If they want to match Red Hat, they don’t just have to download the ISO and sell it. They will have to spend money to get the support team that is required.
I checked your link and was not really impressed. Migrating Windows Server to Oracle linux makes sense because Windows support is even worse. If you can afford several weeks of downtime without too much loss and you have very basic needs, then Windows Server is for you. If you can afford several days of downtime and have more specific needs, then go with Oracle linux. If you can’t afford more than a day of downtime and need a team of experts working for you, then it’s Red Hat.
Edited 2010-11-15 18:25 UTC
It doesn’t matter if you had a bad experience with them. They have good ratings overall but more importantly they have time to improve. That is a key factor here, they are playing a long term game.
No it doesn’t because you are likely to have Windows dependency issues.
Is that how Windows web hosts offer 99.9% uptime? I realize the IT world’s view of Windows Server is stuck in 2001 so your comment really doesn’t surprise me but you should have added a comment about BSOD to make it absolutely clear that you haven’t spent much time with recent versions.
Last I checked AIX had the best uptime, followed by Solaris.
If they improve, then I’m happy with that. I have nothing against Oracle, I use their database because they are the best on the market. I had a good experience with their database. The support is just not on par with Red Hat.
That is what your linked talked about. The business that switched to Oracle linux switched from Windows server, not from Red Hat.
Don’t put words in my mouth I didn’t say. I am not talking about BSOD or bugs or whatever. I’m talking about support. 99.9% uptime is nice but what do you do when you are facing the 0.1% downtime? You call Microsoft? Good luck getting your server back online in less than a week. Their support sucks, that is what I was talking about. They just ignore you. They sell cheap licenses with no useful support. The kind of support they can offer is nice if your server can wait for a week or 2.
Yes, I administer AIX servers as well and I had Sun servers in the past. They have very good uptime and IBM is not bad at support. I’d say they are among the best. I had a contract with IBM where they had to pay us if any downtime was their fault. It didn’t happen though. Note that it is not because it didn’t happen that we could afford the risk to not pay for a good support. The price of support is not a problem at all when your business is worth billions of dollars. The service you get is what is important, the price less so. That is what Red Hat is competing with: IBM. Oracle linux and Microsoft are not competitors yet. Oracle Solaris may be a better competitor than Oracle linux.
In the overwhelming majority of problems that customers call in for, Red Hat will refer you to a “fix” that was released in a newwer version of RHEL than what you’re running. They will demonstrate that they cannot duplicate the problem unless you have a Technical Manager that supports your hardware supplied drivers and RHEL support.
If they are able to duplicate your problem, it will be demonstrated your fault for not being patched up to the latest kernel and rebooting half a dozen times a year.
Now all Oracle or anyone offering RHEL level service has to do is duplicate this. They know it will be weeks to months before you are able to use the recommended version. I think this model is supportable.
Sure they can, they just combine GPL and prop code in the same application, but just like Ubuntu and others distribute ATI/Nvidia closed source drivers, just not at install time. Setup a repo, and distribute it from there? Sure no problem at all.
Well I not talking about proprietary software that is free to download like drivers.
I’m talking about innovative software that adds value to the stack and can be licensed to individuals and companies like Oracle instead of having them just take the source and assimilate it into their offering.
They can’t do that with GPL software either, that’s what the GPL is for.
They can add GPL software to their Linux stack and then tie that stack to proprietary hardware or software.
Small companies are at a disadvantage when it comes to selling open source. Any innovation they come up with just gets assimilated by bigger companies.
And yet Red Hat keeps on trucking. You are not the first one trying to predict their demise, and likely not the last. CentOS was going to kill them, then Novell was going to kill them, now Oracle is going to kill them.
I see a pattern here, wishful thinking on your part is just that, wishful thinking. If it weren’t then all those Linux zelots would have killed Microsoft off with THEIR MINDS ages ago!!
I realize that your whole thesis is that open source is a bad way to do business, and thus you REALLY want Red Hat to fail. But you’re not preaching to your choir here, so unless you have facts to support your notion of Red Hat’s demise then you’re just huffing and puffing.
I never said anything about them being killed off. I just think Oracle has a better long term position.
You still haven’t explained how I am engaging in wishful thinking.
I do get annoyed when open source advocates cite Red Hat as the ideal business model for software when even the CEO has stated that it has limited applicability. Java founder James Gosling recently made similar comments on the limitations of the support model:
http://nighthacks.com/roller/jag/entry/desktop_linux_the_dream_is
I’d rather not see Red Hat fail but instead fork FreeBSD and create their own proprietary Unix that Oracle can’t clone and then heavily fund projects like Samba 4 to create a better Windows Server competitor.
I believe embracing the GPL just leads to a race to zero with big corps like Oracle and IBM eventually coming out on top. Red Hat is making good money now but at some point Oracle will start pushing hybrid software/hardware stacks and Red Hat won’t be able to counter with technological improvements since it all goes into the collective. Perhaps the Red Hat CEO has a plan to deal with this inevitable future but as it stands Oracle has the long term advantage.
Because there’s no indication that Oracle efforts is hampering Red Hat in any way. Which then prompts you to say ‘Yeah now, sure, but you know, I’m like talking about the future…’ which boils down to ‘wishful thinking’.
So Red Hat doesn’t necessarily have to fail in order to agree with your thesis, but just switch to a proprietary model, how very grand of you.
Oracle has a list of customers they have taken from Red Hat.
Red Hat doesn’t have to fail but they should be aware of the threat that Oracle has become.
I don’t think they should switch to a fully proprietary model. I think they should go hybrid so Oracle can’t clone everything they produce.
I wouldn’t rely on that list if I was in your place. Some customers have switched some systems in exchange for press quotes and a sweet heart deal but there are many instances where customers have switched back as well.
Do you have any link to that list by any chance? In the links you provided until then I only found companies who switches from Windows Server.
I think you mean to say that Oracle *could* be the major threat but it certainly *isn’t*. Oracle Linux isn’t exactly the Linux of choice for corporations and enterprises. They talk a good game but pretty much no-one is buying.
Perhaps Oracle is cheaper but are they better? Price isn’t everything.
Edited 2010-11-15 05:13 UTC
Oracle has already siphoned off customers and advertises their RHEL compatibility along with discounted support.
Oracle is the shark in the water, not MS.
On which planet precisely is Red Hat a “small” company? Is it the same one where Unbreakable Linux was a wild success and Larry Ellison was considered to be a nice guy?
LOL! Good advice dude, I’ll remember that the next time I set up a Debian server, I am actually being undermined by Oracle. Where is the FUD button now again?
The point has to do with adding value in terms of code that larger companies can then immediately assimilate into their own offerings that have the advantage of an established brand or tie-in to other products.
It’s a major limitation of embracing open source in business. But I can see few here want to actually discuss this limitation since that might lead to the heretical conclusion that open source does not make sense everywhere and that Stallman is wrong.
Yes, we know Red Hat is the knife in your back since they are the poster boy for contributing to and yet making tons of money off Linux. But why don’t you come back to this if/when there’s actually some data to support your claims. Throwing the ‘Stallman’ card just makes you look desperate.
Red Hat is the knife in my back? LOL no I just find the economics of software to be interesting.
I think the open source ideology that emanates from Stallman actually works against businesses that fund open source since it takes a hardline position against hybrid models.
I believe you fell on the other extreme, thinking that anything that barely looks like free software is automatically against business. This is also an ideology that works against business. Businesses need good software and good service. Neither the Stallman nor the anti-Stallman ideologies matter to businesses.
That must be why I have spoken highly of Nokia and Qt or how I have stated that hybrid models are better. That’s hybrid as in using both proprietary and open source.
There is no such thing as an anti-Stallman ideology. There is no counter movement that has declared open source to be the enemy. Stallman has his critics but there is no counter cult.
They have no legitimate shot with that. It’s miles away from being a competitor to Windows Server.
The only metric that shows Microsoft winning the server war is money. Thats not a shocker given that many servers running Linux were purchased without an OS. If you look at web server stats, Linux blows Windows away 2 to 1.
According to ServerWatch, 76% of companies are planning immediate Linux builds versus 41% planning a Windows Server build.
You are probably talking about that recent study funded by the Linux foundation that only polled the largest companies.
Red Hat doesn’t care about small or medium businesses. Their focus has been on Unix migrations, not Windows Server.
I don’t agree with you often, but yes, the focus for Linux distributors has been on the ‘low hanging fruit’ of Unix migrations.
That can’t go on. A Linux distributor is going to have to look at what they have unemotionally and ask themselves what it will take to create a Windows Server competitor – straightforward file and print sharing, straightforward single sign-on, remote desktop etc. etc. There’s a lot of departmental settings where Windows Server rules and Linux is not even considered. This Unix thing cannot go on forever.
Edited 2010-11-15 17:05 UTC
I agree and would like to see more funding go into Samba 4 to make Linux more competitive with Windows Server.
Oracle is not and never will be a real threat. Its in Oracle’s best interest to keep Red Hat profitable and happy. Oracle has many customers running Oracle on RHEL or their own Oracle Linux. Oracle simply lacks the ability to support these people if anything were to happen to Red Hat.
That doesn’t make any sense. You seem to forget that Oracle has billions of dollars which can fund plenty of engineers. Oracle would love to see Red Hat go down in flames. Red Hat does not have a magical touch that Oracle can never acquire.
Do you really think Oracle bought Sun so they could play nice with companies like Red Hat?
Oracle is just getting started in positioning themselves as a one stop Unix shop. Undermining Red Hat will be one of their main goals.
I think you are under estimating the amount of work that Red Hat does on Linux as a whole. Oracle does not currently have the means to duplicate all that work. Not only that but they have contractual obligations to support existing installs of Oracle DB on Red Hat servers, not to mention their own flavor of Linux. Red Hat going belly up would cause a huge turmoil in their business, and probably get them sued into oblivion. You don’t go out and undermine the databases of the Top 500 corporations without a significant backlash. Remember, Oracle guaranteed to support “Red Hat” Linux on many installs. Can’t really do that if they put them out of business.
And there is basically no one moving from RHEL to Oracle Linux anyway. Red Hat gets almost a 100% renewal rate on there current customers. They must be doing something right.
Edited 2010-11-15 16:29 UTC
No I’m quite aware of how much they contribute to the kernel and even to desktop oriented projects like Gnome.
http://www.linuxpromagazine.com/Online/News/Red-Hat-the-Top-Company…
I understand why some may be defensive of them but I am looking at this from a business perspective. I think there is a very big shark that is circling the boat and it is not Microsoft.
They have billions in cash and some of the top Unix engineers in the world. I’d call that means. Remember that they also don’t need to duplicate work to projects like Gnome.
Oh I don’t think they will go belly up, Oracle just wants to take them down to a minor player.
They have some pretty big customers that have switched like Activison:
http://www.oracle.com/us/technologies/linux/025990.htm
But it is really the long term that gives Oracle an advantage. So far they have only offered discounted support for the same software but they will eventually offer greater value through consolidation and performance gains. They plan on investing heavily into both Solaris and Sparc. Where do you think they are going to be looking for Unix customers? Red Hat will be one of the main targets.
Are you really trying to pass off Oracle’s ‘look at our success stories’ page as some evidence? It’s just as pointless as pointing to the Red Hat equivalent page, it’s advertisment. What’s next, you are going to drag up Microsoft’s ‘Get the facts‘ campaign?
Numbers don’t lie, when Red Hat’s business starts doing bad you will have something to base your argument around. If that happens I will take your statements seriously.
He claimed that “basically no one” had switched.
I presented evidence that contradicted this opinion.
Or are you telling me that list was falsified?
I’ve already stated numerous times in this thread that Red Hat’s business is healthy now but Oracle is the biggest threat to their long term viability.
In the next few years we will hear more reasons from Oracle as to why customers should switch from RHEL. They will aggressively go after Nix support providers before MSFT. Oracle is investing billions into Sparc and Solaris and they will be looking for a payoff to those investments. Do not underestimate Oracle, let’s not forget how the Sun CEO described Ellison as someone who doesn’t get open source. Which company no longer exists and which one has a profitable Linux support business?
Agreed. I don’t know if many realize the distribution model of Red Hat. They have a lot of package maintainers. They do contribute kernel changes to Linux proper as well, but the Linux kernel does not depend on Red Hat coders.
If there is a bug that was fixed in Linux proper kernel source 2.6.24, the Red Hat support team may have no reason to back port a lot of fixes to their distribution purposefully kept at 2.6.18. When they get a customer that has an issue with a particular module, they may incorporate the already made kernel fix to their 2.6.18 channel, releasing RHEL kernel 2.6.18-128.1.23.
Oracle can sustain that and get their own kernel coders should they actually need to contribute anything new to the enterprise distribution. Linux itself will continue on just fine releasing new code from many other sources.
If you look at the stats, Red Hat contributes 12+% of the code in the kernel, the largest percentage by any corporation. Oracle comes in at 3+%. Oracle would have to multiply its kernel development staff by 5 to pick up what Red Hat is currently doing. Thats assuming that the kernel devs actually WANT to work for Oracle. Its not like kernel devs grow on trees.
So 88% of the Linux kernel updates don’t wait on Red Hat. Nice states. kinda proves that the Linux kernel does not depend on Red Hat. The support they offer is not what they contribute for the most part. RH devs are able to back port updates already in newer kernels into their enterprise version. That’s a nice skill, but hardly worth an increasing cost, but it is a very good strategy. No one who signs POs will question it.
It doesn’t matter if there were fewer kernel contributions.
Let’s say Red Hat disappeared and kernel progress slowed…….so what? Linux would still be around and Oracle would offer support to former Red Hat customers. Oracle makes money from support contracts, not kernel progress.