Slashdot Mirror
Oracle Linux?
eldavojohn writes "There have been rumors floating around of Oracle working on their own distribution of Linux. If this is true, it is widely believed that this enterprise edition of Linux would be in direct competition with Red Hat Enterprise Linux. What is spurring the rumors? Well, Oracle chief executive Larry Ellison said, 'I'd like to have a complete stack. We're missing an operating system. You could argue that it makes a lot of sense for us to look at distributing and supporting Linux.' I know that Oracle has been doing a lot more than databases recently, will they go the extra mile and create their own stripped down Linux kernel? If they do, will companies switch to database solutions that are running Oracle only software for the benefits of support and (hopefully) stability?"
Maybe Larry will have them label it Pagoda Linux or Samurai Linux, in honour of his fascination with things japanese
i for one welcome our new samurai penguin overlords!
A feeling of having made the same mistake before: Deja Foobar
I've seen it in a vision.
If this trend continues I wonder how many orgs would be willing to go along for the ride? Imagine a mail server running on Debian, your web server running on Sun Linux, your database server on Oracle Linux, your application server on Red Hat, etc.
All similar but different enough to drive an IT guy batty. Too much of a good thing?
Oracle: We made Linux...expensive!
I didn't know the history of Oracle's head Larry until I saw this the other day. Who knows, he might even have a brother or two named Darrel out there.
Oh You POS
I'm a bit surprised that they're not considering OpenSolaris. Linux is nice, but Oracle has been supporting Sun Solaris for far longer. Using Solaris as their base kernel would allow them to provide a large number of enterprisey (lt;-technical term) features out of the box.
Not to say that 2.6 doesn't have bunches of enterprisey (<-technical term again) features, but Solaris is still a leader in that space.
Javascript + Nintendo DSi = DSiCade
Standby for the usual 'quality' commercial software versus some amateur stuff made in someones bedroom.
davecb5620@gmail.com
This happened (as the FA states) back in April and was covered here at the time.
Just be honest and call it GreedyEgotisticalBastardix.
Seriously, if Oracle did custom-tailor a special Linux distro especially for running the database engine on the x86 platform, especially the 64-bit version, it would be a "Very Good Thing". Already the same exact pile of hardware running Linux instead of Windows 2003, and running the Standard version of Oracle 10g is typically anywhere from 10 to 100 times faster on average for Linux over Windows.
They should know that they won't really have a "complete stack" until they're implementing their own hardware base, so they can provide truly turnkey datacenter solutions. And where did that NC thin client concept go? And here comes Sun with their datacenter-in-a-truck solution.
[
Since we are in Japanese mode, how about Baka Linux?
10 flame warrior experience points and a puff of karma to the first one who figures out why I should be modded down for that suggestion.
Only to idiots, are orders laws.
-- Henning von Tresckow
I'm posting anonymously because I'm an Oracle DBA ('nuff said). Oracle does a make a nice database, but it is hugely bloated for most purposes. And everything else they write is just pure unadulterated crap.
If you look at what it takes to implement their ERP or Pharmaceutical Suite you will realize that they will only ever be a niche player with their own Distro. They write software to require the maximum amount of administration and consulting possible. Their consulting division make a ton of money and they willl never release anything that might endanger that. Also, they have a lot of "faithful" DBA's (like me) that make a really good living keeping the giant house of cards that they call an "application stack" running and recoverable.
Companies with deep pockets will buy it because it's Oracle and pay high salaries to people like me to maintain it all. I'm not complaining, because it's a pretty nice gig, and I might recommend "Oracle Linux" for my company because all the extra crap equals even more job security for the (somewhat scarce) senior level DBA's that have a lot of Linux experience.
With this news, I better get rid of my Redhat stock. This news cannot be good news for ReaHat or any other Linux vendor. I hope I am not too late.
If Oracle began to distribute and support Linux, it would mean good things for Linux in general, while Red Hat deploys Linux to the enterprise sector, they are a Linux based company, whereas Oracle is a much wider known and respected brand, their adoption of Linux for Enterprise could cause a slew of companies to adopt as well.
In a world of acronyms, the words are the real victims.
I don't see anything indicated in his comment that would make me think they are going to create their own brand of linux. My first reaction would be that they would acquire an existing brand. That way they have some kind of name recognition within the linux community with which to start. Then they can tinker/screw with it to their heart's content. I don't know the Linux Distros very well, anyone suggest a name that might be open to being acquired?
"Don't you know you're going to shock the monkey?"- Peter Gabriel
Never. I do not like Oracle. I don't like their DB products and I don't like their business practice(s). Like telling the state of California they needed more Oracle licenses than the state had employees that had computer access for example. I will not use, nor support Oracle....
So, with that in mind...I won't be one of their Linux customers either.
I use Oracle 10g everyday at work. It's a good database and does the job nicely. Unfortunately I have to use another solution made by them: XSQL. I can honestly say that it's probably the worst framework I've ever used in my life. Sure, some parts of it make for rapid development and deployment, but other parts of it are a complete nightmare and sometimes I wonder why they bothered.
Basically I'm wondering why Oracle want to pinch consumers away from Fedora and Ubuntu instead of just working with them to help intergrate their databases more seamlessly into these distros?
Summation 2
What Oracle should do instead is take RHES4U4 or whatever and merely tweak it for oracle performance and release that. I think they'd do well to just get into conjunction with Redhat and make a "RedHat Enterprise Oracle optomized" version of the OS.
Yeah, it would be a subtle fork, but Oracle has enough trouble keeping track of it's DB. I don't think they clearly understand the headache involved in maintaining an operating system.
Yes Francis, the world has gone crazy.
Linux is free but you pay a premium for the Oracle brand name.
With the "building their own linux kernel" business.
If they need to do that, they might. But I don't see anyone in the business world going above and beyond the minimun necessary to sell something.
Twist up their own Debian-based distro and make it sales/support policy to support only theirs, at a fee they feel the market is willing to buy at. Intentionally avoid testing or supporting any other distro and you've got something that's as good as a proprietary OS. Sure, source/patches may be available, but they aren't going to go to any effort to make them workable outside their version. Nor are they obligated to explain to anyone how to compile the whole thing.
I believe an oracle customer would consider making the switch simply because oracle will support it. It will have a gui that MSCE's can learn to use. And oracle has a shiny new certificate to sell too! I don't think the OS being Linux has anything to do with it.
http://www.maxineudall.com/2010/02/should-economists-be-sued-for-malpractice.html
Admittedly, I don't follow or know a whole lot about Oracle, but wouldn't a move like this open the door to them selling a self-contained Oracle Appliance for small- and medium-sized businesses? Of course, they could also supply a list of supported hardware for people to run it on machines purchased elsewhere or built by the company's hardware guru.
This guy's the limit!
Except as a platform to run Oracle on. Oracle doesn't really understand fairness or openness, in large part because its founder doesn't. I'm not saying that they can't figure it out - IBM, after all, went from the most closed of corporations to one of the main sources of energy into commercial open software - but I've always considered IBM to be kind of a special case anyway. Regardless, I have a hard time seeing the industry embrace an Oracle-controlled linux distribution.
It is possible that an acquisition of Novell could bring in enough fresh blood to turn this around... And it would bring in an already-respected Linux distribution.
On the other hand, it makes a whole lot of sense that Oracle would start shipping a Linux LiveCD that runs the Oracle installer, which can be a bitch to get running anyway, and upon which you can run Oracle if you install it to the hard disk. After some time they could switch it to be the only supported platform for Oracle. If you don't want to run it directly on the iron, run it in a virtual machine - although unless you're on ESX or something (whatever it's called now) that's probably going to come with a dramatic performance penalty.
Regardless, it only makes sense for Oracle to provide their own Linux. Why help Redhat? Redhat makes competing products.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Given what I perceive to be their past history and what the industry does regularly, I would expect them to *buy* RedHat, not compete against them. Or, someone else that's ripe for taking (over).
Oracle might be thinking of shifting their application stack to operate the equivilent of embedded. Integrate a custom stripped down, beefed up, optimized Linux or any other open source OS as part of the application install. PXE boot an Oracle app stack directly on open standard harware, or create virtualized guest servers. Takes the OS sysadmin away and moves management to the application level.
Why bother with GPLed code when they could just grab FreeBSD, customize it any way they want, not release the source code and not have to worry about the GPL?
As it all seams to be about capturing audience, why not? This would surely confuse many and weaken RedHat and other competitors.
It's called Solaris. Sure they don't actually *own* Solaris, but Solaris + Oracle go together like bread and butter, as any fule kno.
What the world doesn't need is *another* half-assed Linux distribution.
Larry likes to expand Oracle's reach by purchasing competitors. Consider the world of ERP, where he bought Peoplesoft and JD Edwards. It makes little sense for Oracle to build their own distro to compete head-on with Red Hat. It makes a lot more sense to threaten to build a Linux distro in the hopes of driving down RHAT shares, thus facilitating a takeover. Larry wanted to buy JBOSS, but Red Hat beat him to it. If he buys Red Hat, he gets JBOSS as well. And all of Red Hat's customers. Buying Red Hat would make Oracle the #1 Linux company overnight.
Besides, if Oracle tries to build their own distro, market it via their existing sales channels, and support it via their existing system, Oracle Linux will truly suck. The pricing will be outrageous, the sales process will be the "car dealership" model, and the support will be the offshore model that is not all that great. Oracle makes a great product, but they are their own worst enemy sometimes.
If I were Larry, I would create a great deal of hype about doing my own Linux distro, to soften up the price of Red Hat in anticpiation of a takeover.
If you enjoyed this story, you can also have it in Microsoft, AOL, or Google flavor.
Bah. Pundits. Get a real job.
This makes sense from Oracle's perspective. They want to become a solutions provider, not just a software provider. That's what a lot of enterprise clients want, solutions, not more software to install. When I started as a programmer years ago I never got to see the big picture. Now that I have some experience under my belt and worked with senior managers I can see the bigger picture.
The companies that buy Oracle products want solutions, not new software packages. If Oracle decides to package their own version of linux it would be as part of a package solution. You are no longer buying a database, your buying a data analysis solution. You get a box from Oracle that will do risk analysis, or data warehousing. The database, the OS, app server software, etc, all are just pieces of the solution. Oracle doesn't have a hardware division but the OS would be a step closer to that one package solution.
It is a pain to have to deal with hardware vendors, software vendors, and OS vendors of different stripes. When something goes wrong it's frustrating to see each vendor blame the other for your problem. God knows this happens all the time. By developing an Oracle flavour of linux, they can better provide that solution type package.
I'd switch our platform to oracle/linux, but only so I could get company-paid experience, training, and certification. Then I'd jump ship to something bigger and better. //How I've always done it. ///Make mad $$$
Free childcare classifieds: www.carebrite.com
If ORCL decided to do this I for one would support it. There are so many freaking hoops to jump through to 'correctly install ORCL on linux or solaris or just about any of hte *nixes... Would love to have a preconfigured iso that laid down the OS and was primed and ready for oracle in some fashion... kernel tuned, etc... I know it can be done manually... their whole RAC setup could just totally auto-provision from a single image...
sig goes here!
Larry can certainly try, and Oracle will certainly take up Korea (where they have a monopoly) and a fair amount of the market, but I do not see this working for them long-term. Customers (like me) do not like to get all our software from a single place. The "Lock-in" problem exists almost just as much for Open Source software as it does for proprietary software. What happens if/when Oracle deicded to fork the kernel to better support their Db? What happens if/when those forks start to limit my options? Voila-- I am at the mercy of Oracle for support, compatability, and expandability. No thanks, that was the primary reason we all hated Microsoft, remember?
Red Hat or SuSE work when they come from IBM or HP or Fujitsu because the hardware/OS marriage makes sense. Hardware is pretty much a commodity at this point, so I know I have flexibility. RHEL or SuSE are forced to play nice with all the major HW vendors, so that keeps them honest and in line.
Sorry, but the temptation for Oracle to start dicking with the kernel would be just too tempting, and just too sticky for me.
davejenkins.com |
Pure fanfare.
All Oracle gurus know that the RDBMS is basically an OS by itself aside from 100% H/W control (maybe 99%). A linux distro would only help for packaging so users can have an out of the box DC.
I hope not to start a flamewar but, BSD tends to be more stable than Linux for enterprise purposes (uptime, high load, etc.) even if not by a lot, why wouldn't you choose BSD over Linux for something like this? The other reason I think this would be a good thing is for licensing: They could keep their proprietary tweaks to the BSD architecture as a proprietary edge over other vendors.
Mind you, crusaders, that I am posting this from my Linux-enabled laptop.
This comment is fully compliant with RFC 527.
For example, you can't do a direct SQL query.
Oracle has a strong push towards grid computing and this same business reason applies to Linux. If the customer spends less on hardware and OS licensing, they can afford more for Oracle licensing for a given project. In an example of more traditional Enterprise Oracle on AIX, Solaris or HPUX on a mid-range 4 CPU server, the hardware/OS cost is close to the Oracle licensing ($40K/cpu Oracle Enterprise, ~$150K server/disk). Oracle's goal is to use commodity hardware, cheaper OS and get a larger slice of the pie. In a Linux/Intel RAC cluster the hardware/OS cost is much less and the idea is to achieve the reliability through clustering and the clustering requires more Oracle per cpu licensing.
Regardless of whether it was Linux or OpenSolaris as another poster commented, this would be a **VERY** good thing for people who have to install and maintain an oracle. Especially if Oracle puts a decent updater similar to RedCarpet or RedHat Network. No more fiddling with kernel shared memory parameters, no more worrying about patchsets, no more worring about "if I update the OS, will it break Oracle" (which is the whole point of the OS installation anyway- to support the DB), and no more juggling java versions ( now managed during the install/update). They could just do the "eveything on one disk" software approach, or perhaps they could move into hardware/appliance plug-and-play clustering - just add a node and it configures and integrates.
Oracle did not work well on Red Hat Linux for us, in fact, it worked very badly. I know that raises cackles here and people say it has worked fine for them and give anecdotes about their successful experiences. But this company is in the business of making profits for its stockholders, not promoting free software. Perhaps if we have an expert Oracle/Linux DBAs and a genius sysadmin dedicated to just this and so forth, perhaps it would work, but these types of resources are just not available for us. It has been my experience over the past decade that at most companies, Oracle runs on Solaris. I'd be curious to hear the names of major companies that are running Oracle on Linux in production. Oracle has a long history with Sun and Solaris and has worked most of the kinks out, in my opinion, it is not just their yet for Linux. Perhaps working on their own Linux will help that. Because the only thing those Oracle/Linux developers are going to give a damn about is getting Oracle RDBMS running well on Linux. It might even help Red Hat in the long term.
Anyway, Oracle doesn't need their own distro, they can just ship Andrew's, or Linus's, or Marcello's branch and declare that "The One True Linux". I think people who will spend the $$$ for Oracle would definitely drink that koolaid. But why would you assume a "stripped down" kernel is in the works? If Oracle's going to brand their distro, it would be focused on their toolset, but not necessarily lacking any standard kernel features.
Will businesses switch? Magic 8-ball says, "YES", if they are already running dedicated Oracle servers, and "NO" if they are supporting other apps on the same servers.
Oracle is to Linux as Microsoft is to Open-sourced.
Oracle is to Linux as Military is to Intelligence
Oracle is to Linux as Free Speech is to Facism
Oracle is to Linux as.....(fill in your own here)
Why do I feel like I am taking the SAT's, GRE's, etc. all over again?
This just makes no sense. Why would a company develop an operating system that is a rebranded version of one that is available for free. Also, the way I understand the GNU, wouldn't they have to release the changes they make to the kernel to the community at large? In that case, why pay for it when you can re-compile it yourself. If not, and they pay to market a proprietary version of it, you know the cost of owning a proprietary version will far outweigh the benefit of marketing it. In terms of supporting only "their" version, this could backfire as I could see some of their client base saying, "Okay, in that case, we'll take a look at some alternative solutions."
Just another nameless binary in a crowd of 1's and 0's
IF this is true, the only reason I can see Oracle doing something like this is if they integrated their database with the OS so they became one. Which probably isn't too good of an idea.
Expect the first revision to kind of suck.
m &q=l&c=RHAT
http://finance.yahoo.com/q/bc?t=6m&s=ORCL&l=on&z=
Suppose Oracle supports their own, reduced, version of Linux (with any performance enhancements that they deem necessary). If they "partnered" with a hardware vendor, you'd have a single stop for your database server needs.
You'd get your BIOS updates, OS updates and database updates from a single company that could afford to do the testing so the load on your IT department would be reduced.
You could even order it in a cluster configuration.
But what good is a database server on its own? With a bit more work, you'd be able to buy a webserver box (hardware, OS, Apache, etc) pre-configured to hook into the database server they sold you.
From Oracle's point of view, this would be a great way to get even more of the market and to stop any gains from MySQL or others.
From the corporations' point of view, this would be a great way to reduce IT costs by reducing the load on your internal IT department.
If Oracle does it right, they'd even be able to offer you dial-on-demand DBA services for their products. Why pay 6 figures to hire an Oracle DBA when you can pay 5 figures for a DBA service contract with Oracle?
I just got back from the results meeting for an Oracle RAC evaluation, done on Solaris 10, with various filesystem options. The O/S support was no big deal, but the DBAs had a really rough time of it, even with Oracle, Sun, and Veritas consultants along for the entire ride. One DBA checked out halfway through the three-month process, and the replacement couldn't go a single day without mentioning how simple RAC setup and configuration is with Red Hat. The project is on hold, pending better explanations from Oracle as to why it took so long just to test it.
Even though Oracle and Sun seem to go together--at least, in the minds of IT management--there are several documented gaps, e.g., Oracle on Solaris with NAS storage is painfully slow without tuning. Fortunately, there's a whitepaper that covers this particular case pretty well. Not so for every deficiency in the Oracle/Solaris combination.
Surely Oracle/Linux has quirks of its own, but like the ubiquitous Oracle/Solaris environment, these will surface with time + a growing installed base. The firm I'm supporting is very risk-averse, e.g., many meetings must precede even the evaulation of new technology, with many more meetings (and man-hours) before the pilot. I don't see them even testing Oracle on Linux for another year.
By the same token, how many early adopters for a DBMS (even one with ludicrously huge market share) with its own O/S?
See also the Pick Operating System.
"Press to test."
(click)
"Release to detonate."
small correction: Oracle bought Peoplesoft. Peoplesoft had already bought JD edwards.
One
Raging
Asshole
Called
Larry
Ellison
When Oracle says they're missing an OS, I don't see Oracle 'making' the OS. I see Oracle buying Novell, or Redhat, or maybe one of the more marginal linux distros. Who actually believes that Oracle would build an OS from 'the ground up' when they can buy the expertise from an entrenched distro...
Bye!
This seems like an move for Oracle to be able to offer a "database appliance": prebuild boxes with Linux and Oracle, and sell those to PHBs as "drop in" solutions. Many in management would fall for it.
Oracle is a database. What separates them from the rest is that they do thing correctly. They fix errors even in documentation quickly, and make things works "as they should". Having been an Oracle DBA, i loved it, and i do not know of a comparable RDBMS support structure.
So, even if they made ice cream, or pocket-protector protectors, i'd have to take a look.
In Linux, i use Debian. They also try to do thing correctly, though they have their pitfalls. I'm a bit suprised Oralce didn't choose Debian, but i'd have to guess it'd be similar to it, just not so open to packages.
Have you read my journal today?
I doubt it is terribly bad news for Red Hat. Even if Oracle create their own distro I doubt they would get away with ceasing to certifying their products for any other Linux distros. There are simply to many people with already established contractually sealed working relationships with SUSE and Red Hat. Of course Oracle will recommend the use of Oracle Linux® (can they even register that as a trade mark if it contains the word 'Linux'?) and all updates to Oracle products will appear for Oracle Linux® first and only 2-3 weeks later for SUSE, Red Hat etc. but that's about it.... unless Larry wants to piss off every last one of the sizable number of IT professionals world wide that haven't been assimilated into the Microsoft collective yet. Even if they do stop certifying Oracle products for anything but Oracle Linux® I doubt it would be much of a problem to get Oracle products working on un-certified Linux distros. It would simply take a bit of debugging and howto files for Oracle instalations on un-certified Linuxes are easy to find.
Only to idiots, are orders laws.
-- Henning von Tresckow
Would you use an optimized Oracle linux to host a Postgresql db?
What if the database replaced the filesystem, i.e. Oracle *is* the operating system? All of your files are just B/CLOBs, all apps are packages called by the Oracle db, and so forth. I don't think I'm reaching (much) on this. Although I do think Bluetooth support will kill it.
-BA
I'd say Oracle already does something similar, by exercising a stringent set of hardware and software requirements. One of our groups here has been locked into Oracle for the last dozen years. Each major revision of Oracle came with new requirements of "Oracle only supports X hardware running the Y operating system". There are several options for X and Y, but it obsoleted our systems EVERY TIME. The whole monstrous process was driven by the need to have Oracle support. Now that group is reevaluating its needs, and starting to build a demo system using Linux and MySQL. If it takes off, Oracle will be out the door.
As much a fan of Linux as I am, I would dearly love to be able to dtrace to track down performance bottlenecks on my Oracle server.
sqlplus sysdba@myserver/tiger
update SYSV_INIT.INITLEVEL='6';
commit;
(or something like that. I'm a SysAdmin damnit, not a DBA)
What if it is just turtles all the way down?
Oracle would make a Linux distribution so they could provide a one-stop 'solution'. Oracle targets the 'Enterprise' market, and most enterprise IT admins want solutions and support. They do not want to install another piece of software and fix it whenever an OS update breaks the software; they also never want to hear the phrase 'I'm sorry, but that configuration is not supported.' Controlling the OS and the application (and maybe even the hardware) would allow Oracle to focus on providing and maintaining a drop-in solution (think: Macintosh) for databases instead of supporting all configurations poorly (think: Windows). Most IT admins DO NOT want to patch and recompile the kernel of a mission-critical system (risking possibly hundreds of thousands of dollars in damages because of lost uptime). Even though all the code would be open-source, to compete with OracleOS in the Oracle-ready database OS market, an OS vendor would have to provide both top-notch technical support AND test EVERY new version of Oracle for problems. Oracle would have the advantage of mind-share AND would be in control of the entire stack. If an Oracle competitor ever emerges, it will probably not be able to overtake Oracle's distribution. Why did TerraSoft never overtake Apple?
Oracle could work with an ISP and mail out a CD or DVD with the OS, allowing home owners to try it ala AOL. While I am sure that they are thinking of a server space, this would get users AND semi-geeks use to it. While AOL is derided here, this approach apparently worked for them. The time for Oracle to push this is while they have a good name. Otherwise, they will become like Aol, Novell, or Wordperfect and then be unable to do this.
I prefer the "u" in honour as it seems to be missing these days.
... for their network appliance systems.
And Larry mentioned NetBSD for an Oracle database appliance a long time ago.
But, with their recent Linux focus, any NetBSD bias probably evaporated long ago. Linux is flashy, and because of this it is easier to find admins and apps.
If Oracle can deliver a stripped down, optimized version of Linux or, specifically, Red Hat, then they can simply deliver the instructions to configure a stock system. Then any SA can configure, document and maintain the system (and/or similar system), rather than a special "Oracle Linux SA".
It must have been something you assimilated. . . .
Last time I installed Oracle on Windows it installed Perl, Apache, Java.. I forgot whether I just joked about it, but in my mindset it even required Cygwin somewhere in all those gigabytes. Shipping a customised kernel and filesystem to run seem like a step consistent with that.
Oranux 12j (Release 10.3.0.1.2.0.4.1.3). . .?
What?
/usr/bin/fortune is my oracle :)
You see, the people who BUY oracle are not in the vast majority of cases the people who USE oracle. Mixing magic buzzword pixie dust into oracle lets some C-level bumblefuck put on his resume that he "bravely integrated cutting edge solutions with the whitespace of synergistic proven data management technolgies in the enterprise to improve our core competencies and effiencies for increased shareholder value" or similar. Linux, AJAX, Ruby on Rails, anything new and Shiny will do. Journalists for eWEEK or whatever other industry rag will write glowing articles, POs will be written, and a great round of back-patting will ensue. Somewhere a poor bastard will have to figure out how to actually manage/integrate/use the terrible bloat to do something useful for the business, but fuck him anyway, he's expendable.
News for Geeks in Austin, TX
Why not BSD?
/etc/rc) is really a pain.
1) Lackluster commercial support - Linux tends to have better hardware support, drivers, etc.
2) SMP support on the *BSDs is still young and immature. Linux, in comparison, is quite mature, and does very well on an 8-way system. BSD *might* do it, but much beyond 4-way is a sail into uncharted waters. I'm already running a cluster of 4-way boxen, so 8-way or more is not very far off, given our company's annual 2x growth curve.
3) "It's different". Yeah, it's very similar, but if you're already used to the "Linux" way, having to rediscover how services get initialized (a la
4) Linux is "good enough". It's obvious that whatever metric is needed to be able to be "enterprise ready", Linux has passed it. Granted, nobody agrees on what that standard is, but most people agree that Linux can do it.
I have no problem with your religion until you decide it's reason to deprive others of the truth.
Somehow I doubt that they would bother modifying the kernel. That sounds like an invitation to trouble.
Probably, they'd toss together a fairly normal distribution with stripped down support for application software and and easy installer for the Oracle database.
Going the extra mile, knowing that they have some rather large government clients, they might also try to implement the fixes for any IAVAs that they can by default.
For a large IT department this is going to make a lot of sense.
Oracle will provide a 4u box that is an all in solution. Plug in and go. DB Server in a box.
Oracle ERP could be shipped the same way.
Now what if other vendors came onboard with this.
For any large package setting up the server and installing, configuring the software is a nightmare. But I'll buy an overpriced box from you if you configure everything before it ships. Oh, and I'll pay through the nose for you to support it. No my staff won't touch the configurations just add and remove users from one of five or so classes.
I imagine Dell will be look for partners shortly. Rebadge this Dell as an SAP server. You want your Dell/Siebel. How about Dell/Cognos.
Setting up a new company I'll take a Dell/SAP, Dell/MailBox, Dell/Oracle, Dell/Archiver. Default settings.
It won't be long before HP, IBM et al would need to get onboard.
Now I've got my data center under control and cost effective. What about these fscking PCs on people's desk. Oh you you have an IBM/Productivity, install it in the data center and allow network access via dumb termninals. Costs go down and control goes up. I'll have two please. Yep and the support contract. We'll start with the software client but next PC upgrade we'll go with the network displays.
Personally, this is could be a big win for Oracle if Larry doesn't fsck it up with his ego. Keep Sun out of the picture and start adding items to your product list.
Oh, I need an Oracle/IM box and an Oracle/WorkFlow and an Oracle/DocStore. I just have to plug em in right???
There is going to be a huge market when companies start treating software this way.
The only way MS is going to compete is to do the same, that'll be fun watching them support there own software.
They should name it "Lorax".
Oracle can't even properly beta test its database before releasing the horrid mess to the world (10g RAC). I can only imagine the unmitigated disaster that awaits an Oracle operating system and database combination. It would be much smarter for Oracle and Red Hat to offer preinstalled and preconfigured boxes based on the customer's intended use, fully tested and certified to function.
Not another Linux ............. Please.........
The world doesn't need it. Instead of unifying all linux-related activities
to get *one* good OS, everyone with a 'feature' or two in mind spawns a
new distro...
FreeBSD rules! Get used to it.
This seems like a nice idea on the surface. Each software vendor certifies their software on their own software stack, all the way down to the OS. You can order a box to do a task, and be assured it will arrive and work. It will have lower support costs on the ISVs side, since they know what they are talking about. These are good things.
However, in the real world, we don't support isolated boxes. My database servers are required to integrate with our LDAP directory and security architecture properly. RedHat and Debian already do this differently. Do I need to learn two configurations, or more than two, one for each ISV's Linux flavor we deal with, and how to get them all to work together? This is complicated by the fact that not all of these different distros are likely to even support these things. Sure, RedHat comes with libpam_krb5, and so does Debian... but will Oracle Linux? Will it come with the same version, or will it support Heimdal instead of MIT? Oracle has to test this too... but not just inside Oracle.
Each OS stack will need to be tested with each other OS stack to assure compatibility. So now we've gone from Oracle having to test *their database* against RedHat/Suse/Debian/Ubuntu, and changed it to Oracle having to test *software they know nothing about* against RedHat/Suse/Debian/Ubuntu/*AND 50 OTHER ISV LINUX CONFIGURATIONS*. This is not an overall improvement, not for Oracle nor for an IT shop.
The number of Linux distros actually used in enterprises is actually pretty small. RedHat/Suse/Debian/Ubuntu. There really aren't significant figures of much else. Oracle has it nice in that they only have to test *one product* against *four*.
What was funny is all the vendors were in a room an agreed we had a supported configuration until the trouble happened. When the Sh$t hit the fan basically you started going in a loop. To get a supported config for one vendor you invalidated it for another. This was all on a 24x7 site...you can't take the site down constantly, and switching to backup databases is not a trivial matter.
The thing is in the end CEO's don't care about anything other than it doesn't work. You spend lots of money on the Oracle license and that darn database is crashing all the time. Stability is everything in an enterprise environment. Also as an enterprise customer I do not want to talk to 4 different vendors, I want just one (easier to choke one throat!)
I see it already. Oracle can create the 10g line of everything. Oracle Linux10g Oracle DB 10g why not Oracle Messenger 10g and the 10g framework and what the heck, their new web portal Live 10g! And then when customers ask what 10g is, they find out it was just a cooky marketing term applied to everything the company created that year! MS thought it was a good idea!
Oracle 6 used to run on everything, included the kitchen sink.
Then as the RDBMS grew into Oracle7 it was just a bunch of selected platforms (too hard to keep track of all platform nuances and quirks).
Then Oracle Parallel Server came in, and it was a mess with the DLM layer, some implementations in userspace, others in kernel space...
Then it was Oracle8, and a generic DLM was written, and the implementation variations greatly reduced...
but in 8i it was fairly clear that interactions between DLM and the clusterware layer were the new mess.
With 9i RAC, a couple of platforms (Windows and Linux) had the Oracle Cluster Manager (oracm), but it was a stopgap solution while the acquisition of the Digital clustering technology was turned into a product.
In 10g, a generic clustering solution appeared (CRS), so now Oracle RDBMS runs on top of Oracle DLM which runs on top of Oracle CRS - on any supported platform.
The missing step would be IMO to have a Linux distro which, yes, runs on x86-64, POWER, SPARC and Itanium, to cover all hardware on which you currently would run Linux, Windows, HP-UX, AIX and Solaris... get hardware from any vendor - barebones hardware, no "proprietary" OS ! - and Oracle Linux on top of it.
As I mentioned back when this came up in April ...
Oracle already owns their own distro... home wiki google
--
Slashcode bug # 497457 - unfixed since December 2001 - Go look it up!
o/~ Join us now and share the software
Didn't they try and failed this idea with Raw Iron before?
I agree, and this is exactly what I said in April, when this whole rumor started. But there's not much evidence that Larry is really interested in such a take-over. If you look at TFA, they're just dredging up this quote from the Financial Times from back then. There's nothing new from Oracle on this. If anything, Larry has refuted the idea of buying up either Red Hat or Novell, repeatedly. Just a few days after this rumor started, he reiterated to the Financial Times his belief that Red Hat is an unsuitable purchase because "they own nothing." Still, he likes to drop hints and innuendo about things like this from time to time because it creates buzz around his company and that's good for his own stock price. If the little voices are whispering rumors again now, I assume it's just because of the Oracle OpenWorld show that kicks off in San Francisco next week.
Breakfast served all day!
If Oracle makes a Linux distribution, it would be used ENTIRELY for the purposes of providing a system for the purpose of running the Oracle database. Oracle is a software app that is targeted at large enterprises; there is very little reason for a home-user to use it.
---- "XML is like violence. If it doesn't fix the problem, you aren't using enough."
I used to work for Sun, and a while back Oracle had an initiative called "raw iron" which was meant to be a single product, the Solaris kernel with Oracle directly on top. A single vertical product that was installed directly on the computer.
I think that perhaps now, with the arrival of OpenSolaris, its time may have come.
It was called Raw Iron, and was introduced in 1998. It went over like a lead balloon. People said they wanted an appliance, but they also wanted to tinker with the OS. Just search Google: "raw iron" oracle and you'll see what I mean. And for further proof: http://asktom.oracle.com/pls/ask/f?p=4950:8:148128 41480014660872::NO::F4950_P8_DISPLAYID,F4950_P8_CR ITERIA:7931107631402
I was operating under the assumption that Oracle had cross-licensing agreements with Sun for access to each others tech. I guess I also erroneously assumned that msoft charged something from Oracle to enable Oracle software to work with windows. I guess I'm really out of touch...
Sorry for wasting your time.
D
Previously: "Linux... Toward the Sunrise..." Now: "Linux... Toward the-- No, now, part of Every Sunrise"
Waste time? Nah, it was the highlight of the thread! ;) :P
Javascript + Nintendo DSi = DSiCade
This makes sense to me. Folks installing Oracle don't really want to care about the OS that it's running on... they're far more insterested in making sure that it just runs and runs forever. Actually, I've been wondering when Oracle would do this. The database guys are one of the few people who can still benefit from the extra bit of customization that owning the whole stack can bring. IMHO, most other apps would be better off shipping a VMWare image.
The biggest dent into all of this will come to RHAT. Depending on their current agreement with Oracle, Maybe they should start opening up solutions and platforms based on running MySQL, PostGreSQL and such and tye it together with some goodies made with JBOSS.
I don't know if this is humorous or a serious business proposal.
But this is one of the things in the middle and upper middle size tiers that is lacking
Oh yeah, you're those linux guys, now what can we do with it.
Go on the offensive. Being defensive isn't working. We have this product that is scalable that can handle all of your needs and then some. Provide them with a solution. You've got a lot of tech's convinced but It's like a Mac and PC commercial I can do this, but so can I, but I can do it better after I tweak a few things here and there.
RedHat needs to work on training people on providing solutions with any part of their systems. Using PHP, Mail, perl, python, show them how it can be done, not just how to keep it running.
Maybe they should advertise their classes as a place where you send those technicians that don't do much of anything but run around scared and installing patches. Send them to us, we will sprinkle them with a special fairy dust for a couple of weeks and when they come back they'll be able to make things that solve your communication problems between marketing, production, and sales.
There's plenty of work to be done in those areas. Especially in the mid to upper mid size companies. (100 to 2000 employees).
Stretch it out show them the way.
Matthew, Ron give me a call I have some ideas here.
He who said 1,000,000 monkeys on 1,000,000 typewriters would eventually type the great novel, never saw an AOL chat room
They could also choose a BSD base. I believe Apple and Sun are also using BSD, which seems a natural fit for a company like Oracle.
--
Don't like it? Respond with words, not karma.
It is a logical progression. They are already doing something similar with their ocfs, which they recommend for 10g clustered db's on Linux. You can download and play with ocfs all you want, just don't ask for help unless you have a db license.
Keep in mind that Oracle only "supports" Red Hat as far as the kernel. And only if you do not modify your kernel. They support nothing at all in userland outside of their own software. They don't even support lvm. And by "support" I mean they will answer your technical questions and try to help you troubleshoot kernel issues, but it does not mean they are going to distribute patches to the Red Hat kernels.
However if they do maintain their own kernel, then perhaps it would get patched as quickly as their db software. Plus, I would feel better running ocfs if I could be certain that Oracle really knows what it's doing with Linux.
"Crude and slow, clansman. Your attack was no better than that of a clumsy child."
Your politeness about it. It's one of the highlights of the day for me. (I have to begin looking for a new contract after this Friday... assignment ending 2.5 months earlier than I'd hoped for... but such is life...)
Regards,
David
(hehehe, Slash image word: atheism... this machine must REALLY have a sense of humor behind it...)
Previously: "Linux... Toward the Sunrise..." Now: "Linux... Toward the-- No, now, part of Every Sunrise"
...Oralinux
... flying chairs.
yes, we have no bananas
This was my thought exactly. They'd probably restrict their support to a subset of a few well known hardware vendors' products.
Oracle's support is already less then perfect, especially with Linux. I wonder how it will be when it gets bombarded with the questions that are now fielded by RedHat and SuSe.
On the other hand perhaps this could herald some improvements in Oracle's support for Linux. Now they have guys who apparently have never heard of Linux bonding and 8021q. It seems interfaces names "bond0.1001" and "vlan1001" don't work with RAC because they're too long.
I'm sorry if I haven't offended anyone
Red Hat claim that they cannot give out trial copies or dev copies of RHN Satellite because it uses Oracle. They won't even give you a free dev license if you're paying over 100 grand a year for other products, so they force you to pay for Satellite twice if you want to, you know, TEST changes before applying them to live. The worst is that our site license for Oracle gave us FREE dev licenses, so the reasoning didn't really apply.
:)
The main reason they gave for using Oracle instead of PostgreSQL and dropping their Red Hat database was their partnership with Oracle. Because they have such a special relationship with Oracle, they couldn't in good conscience compete with them.
I really, really hope they get screwed by Oracle seeing as they use Oracle to screw people
Well, sort of. The Japanese distribution Miracle Linux is partly (50%) owned by Oracle. Miracle Linux is also a part of Asianux, which sort of looks like Windows XP, by the way.
SunOS 4.x was the last of the BSD based.
SunOS 5.x (Solaris 2 and up) are based on SVR4.
All the more, then. From the wikipedia link: "Popular SysV derivatives include Dell SVR4 and Bull SVR4. ...Sun Microsystems' Solaris Operating System and SCO UnixWare, both based on System V Release 4."
Those sound like the kind of names I expect to hear Oracle wishing to have associated with their name. Well, perhaps not the SCO name...
--
Don't like it? Respond with words, not karma.