Oracle accused of breaking US competition law over Solaris support Oracle has been accused of unfair competition and of breaking US anti-trust laws over its Solaris support business. The claims are made in a counter-lawsuit lodged by Solaris fix-it company Terix, which had previously been dragged into court by Oracle for allegedly stealing the database giant's copyrighted code. The Terix …
"But what about houses that were running Solaris kit before Oracle took over and changed the rules?"
The same thing that would happen to any open source project that was abandoned. The companies using the open source version have the option to continue to develop fixes and patches on their own using the code they have the right to use (if they have the in house skills that is), but they have absolutely no right to the development of patches and fixes done by a company they have no association with at all.
Seems like a clear case to me. You buy a commercial Unix from Oracle and you get fixes for software faults from Oracle.
Yep clear as day, buy a Ford and you're only allowed to take it to Ford garages, buy a GM and you're only allowed to take it to GM garages. Buy a House and you're only allowed to have it maintained by the original builder.
I wonder who Oracle use to service their AC systems? I wonder if it's the OEM? I wonder who Oracle get to operate and maintain the coffee vending machines they must have in their offices, I wonder if it's the OEM? I wonder who Oracle get to... etc. etc.
This is a nasty tactic many in this industry are now trying on... trying to lock those who brought their product into paying them money every year irrelevant of service delivery standards, and competition laws. I'm hoping sooner or later the competition authorities will catch on and fine one of them the 10%, so the message is delivered in a very clear way.
"This is a nasty tactic many in this industry are now trying on... trying to lock those who brought their product into paying them money every year irrelevant of service delivery standards,"
Actually I wouldn't have a major problem with that - IF it is accompanied by a requirement (enforceable in law) to maintain and support the software in question for the term of it's usable life. Not it's planned life according to the vendor, it's usable life..
End of life for an O/S - no thanks, I paid for it and wasn't told (at time of purchase) the date on which it would be disowned by those happy to take that money, Alternatively, End of life for an O/S - fine, just release into the public domain all of the information required for others to support it.
Personally I would rather see software covered by rules that say defects must be fixed regardless of time since release - on the basis that if you accept money for something, it should work as advertised and if things like security holes aren't documented at time of purchase then they shouldn't be there - in keeping with that view, I'm just off to guide that squadron of pigs into a safe landing.
Its called quality control. You buy a license to run the OS and if you need support, you get it from the licensee. Nothing to do with lock-in, everything to do with providing an enterprise level service and the only way you can guarantee that is to control the service. Do you think RedHAT would be happy if SUSE would be giving support for RedHAT licenses? I also recall HP having a problem with 3rd party support providers providing firmware updates on HP systems. Similar situation.
Its called quality control. You buy a license to run the OS and if you need support, you get it from the licensee. Nothing to do with lock-in, everything to do with providing an enterprise level service and the only way you can guarantee that is to control the service. Do you think RedHAT would be happy if SUSE would be giving support for RedHAT licenses? I also recall HP having a problem with 3rd party support providers providing firmware updates on HP systems. Similar situation.
I loved the provide enterprise level service bit of that statement, if they were providing enterprise level service to their customers at a level the enterprise considered worthy, they wouldn't have any competition to begin with.
No it's not, it's a deliberate tactic designed to try and lock customers in.
What they've done is build themselves a market by selling their product as enterprise ready, that it's the kind of product you can bet your company on. Then after the fact decided that now you're onboard they're entitled to milk you.
That wasn't a problem as long as you had the option to take your business elsewhere if/when you weren't happy with the service you were receiving for the price you were paying.
They don't like that model, so they've decided to change it, to a model where you WILL pay them, or you'll run the software they sold you as 'enterprise ready' and safe enough to bet your business on, in an exposed and/or unstable state. It's the "pay up or else" maintenance business model.
Oracle aren't alone in thinking they can deliver such "pay up or else" terms and conditions to their customers.
It has nothing to do with quality control, if you go to a third party for support you do not blame the OS maker when the support is crap. You blame your support contractor. In no way would it affect the quality control of their enterprise solutions. In fact, if they did offer by far the best support then the market would choose to buy support from them anyway.
"Do you think RedHAT would be happy if SUSE would be giving support for RedHAT licenses?"
Funny you mention that example. Oracle does provide third party support for what is essentially Red Hat. They are not pleased, but they can't stop them.
If they want to be a proprietary software company, that is fine. When they purchase a company committed to open source, Sun, and then decide to roll it back into their proprietary system, it is going to rightfully upset people who have had the game changed on them post-purchase.
Although I do understand why companies would not want third party support for quality control reasons. When everything goes wrong, which it might, they go back to the manufacturer and try to get support even though they haven't paid for it.
Although I do understand why companies would not want third party support for quality control reasons. When everything goes wrong, which it might, they go back to the manufacturer and try to get support even though they haven't paid for it.
In which case the manufacturer will either refuse support, or invoice the customer. During my time as a FS tech, I've had calls (handful each year) where the customer had a 3rd-party service contract on some piece of kit, and there was a problem they couldn't fix. They would have already done the "please come, we'll pay" song and dance with contract management (the ones responsible for finances in those matters, even though the customer might not have a contract with us in the first place), and I had to verify that they had done so, plus then get a signature on the time/materials sheet. But apart from that, it was no big deal, and not at all unusual. Same with software.
True, it is fairly common to reinstate support or pay for time and materials. It is still a hassle though. With most OEMs, they go to the back of the queue, after everyone who was already on support, which usually creates a problem because they wouldn't be calling if it wasn't on fire. Then you get into these situations where they say... "we spend $x on other products, fix this now or we will never buy anything from company x again." These situations always are money losing propositions for the OEM too. Support contracts are set up to spread the high costs of these crisis situations across multiple customers. If a customer only calls when they need a team of engineers and says that now they are willing to pay $800 to put the server on maintenance, the model isn't going to work. Similar to only purchasing car insurance after you have been in a car accident.
It also has a negative impact on the OEMs reputation within the company and elsewhere, because the IT infrastructure group, generally, just tells everyong that Oracle, IBM, EMC, Cisco etc "crashed"... failing to mention that they weren't up to the current support levels, off-maintenance, etc.... Third-party support can work, but it takes a very fair minded customer.
Agreed, it is getting ridiculous the cost of support contracts and what little support you actually get.
Companies are often charging out their basic support reps for £1000 a day (and it will always take a few days, with testing, backup staging, live deployment etc just to do simple point updates).
These costs are more than consultant doctors get for day's locum work (and they have had far more training, education, knowledge and greater risk associated with their work.)
There is no competition to who can provide these services and sometimes they are increased 10% or more per year. As soon as they become too expensive to attract new business the existing customers get stung even more to maintain their profit line.
Unfortunately the costs of support just become another opex that is rolled into the cost of doing business for the biggest companies when in reality it needs the majority of customers to fight back against it. However the suppliers know that you are being held to ransom because very few if any can abandon ship and go with a different provider.
I think it would be great to have a serious enquiry into allowing third parties access to support patches and updates and provide competitive services across the industry (as long as the customer has rights to those patches in their own support contract) or the ability to buy individual support patches and fixes for a reasonable limited fee.
Even the big, bad, money-grabbing Microsoft allows customers free access to patches and small updates and would not care if a third party installed them.
Nonsense! Solaris support is included in Oracle Premiere Support for a fixed/annual rate of either 8% (OS only) or 12% (OS plus HW) of net HW price. A $10K server will cost just $1200/yr for life of server. And Oracle's premiere support includes many things like 24x7 2hr response time support for hardware, 24x7 support for OS and virtualization (Oracle VM and Solaris Zones), OS, virtualization and management software licenses, Management HW, SW support and Proactive Support. Considerably cheaper than RedHat/VMWare/AIX/etc out there. http://www.oracle.com/us/support/premier/servers-storage/overview/index.html
If you had followed the OS scene, you would have known that Solaris is the hottest OS right now. It scales far better than any other OS, with 1000s of threads. AIX does not scale that much, the largest P795 has 32 sockets and 512 threads and AIX was rewritten to handle that many threads. In 2015 Oracle will release SPARC servers with 16.384 threads and 64TB RAM. SPARC T5 is much faster than POWER7+.
Regarding Solaris, everyone is copying or porting tech such as Containers, ZFS, DTrace, SMF, Crossbow, etc. Let us take DTrace for instance.
-IBM AIX has copied DTrace and calls it Probevue
-FreeBSD has ported it
-Linux has copied it and calls it Systemtap
-QNX has ported it
-VMware has copied it and calls it vProbe
-Mac OS X has ported it
-NetApp engineers discussed it on blogs and talked about porting/copying it. As NetApp is based on FreeBSD, they have it now.
Every major Unix OS has it now. Does anyone copy anything from AIX or from Linux? No. Everybody wants ZFS or copies it (BTRFS, ZFS-On-Linux, Apple considered porting it, etc). SMF is copied under the name systemd. Crossbow is copied under name Open vSwitch. Solaris Containers are copied by Linux too and by AIX (called WPAR). There are enterprise companies selling big ZFS storage servers. Everybody is watching Solaris and copies stuff from Solaris.
I dont see anyone talk about tech in AIX or wanting to copy something from AIX? Or from Linux? What tech does Linux have, that other OSes wants to copy? Nothing. Linux is a Unix copy, and everything Linux does, is copying from others.
And besides, AIX is going to be killed off says IBM. The new POWER8 servers will focus on Linux, not on AIX. AIX has no cool tech and is not developed heavily anymore. IBM is betting on Linux. Read it yourself:
http://news.cnet.com/2100-1001-982512.html
"...The day is approaching when Linux will likely replace IBM's version of Unix, the company's top software executive said,..."
AIX will be killed off. The hottest OS that everone watches today, is Solaris. Your post is just ignorant, because you dont know that everyone talks about ZFS, DTrace, etc. You should follow the OS scene better.
@MadMike - hmmm, all those things you list, weren't they all developed by the old Sun? Has Oracle really brought any major new tech to the table since the purchase? Or are they just milking the cow until it dies?
And "hot" implies "fashionable". So your claim that Solaris is the current "fashionable" OS means it's the Izod and penny-loafer of the OS world. Let me think - when was the last time I saw someone wearing an Izod shirt...
Fanboi much?
I'm an IT contractor in the financial industry and over the past decade I've put in very few NEW Solaris installations and the majority of the work I’ve done is port products and infrastructure to RHEL or SUSE. Its cheaper, easier and you get choices you simply can’t get from Oracle. Take a look at job boards and you see less and less requirement for Solaris engineers, thats because less and less people are using it.
DTrace, ZFS, containers are all great technologies, technologies that Sun started, not Oracle and I’m glad that they’ve been ported or copied to other OS’s. The open licensing by Sun is the reason they’ve been ported and improved upon for other OS’s in much the same way Solaris uses the Gnome desktop for one of its UI’s, and Apache for its web server etc. Great things are developed everywhere and ported everywhere so long as the licensing allows it. I’ve yet to see a NEW great product be developed under Oracle though, let alone a great product with an open license. Oracle tend to take and not give.
SPARC is a nice architecture, but the fact is, it’s too expensive to maintain SPARC development when COTS x86 will work the majority of the time and businesses can let Intel and AMD do the development (or ARM in the distant future!).
Solaris itself is a dying OS. Ever since Oracle got hold of Sun businesses have been moving to Linux en-mass, their support is terrible and not being able to go through third parties has caused headaches for its end users. The OS has a few nice features, but these features are things that are used by 2% of its users and most of the time an alternative solution can be found.
The reason IBM are putting more support into Linux is because they are a smart vendor who can see whats coming (or already hear depending on how you look at it)
Firstly, I love Solaris – I love IRIX, BeOS and host of other OS’s... but if you are putting your career on Solaris over Linux then your being extremely naive and letting your emotions rule and may find yourself in a very niche career with few openings (that may be good OR bad for your salary!).
"I've put in very few NEW Solaris installations"
+1 - havnt seen anyone putting in new Solaris kit in years - only upgrades for existing.
"and the majority of the work I’ve done is port products and infrastructure to RHEL or SUSE"
Ditto, but in my quite wide ranging experience, most enterprises go Wintel first for midrange stuff and only stick with UNIX if they really need to - for instance as the preferred platform for Oracle DB.
Solaris is a good OS... it was the best and others had to catch up.... the problem is that since Oracle purchased SUN, Solaris has been slipping. It is not the "Hottest" OS... not even close.
You mention the single-instance scalability of the OS.... even if I simply agree 100% with you... I have to ask, who actually scales anything to that size in the IT industry? The answer, I am afraid, is very, very, very few customers.
What I will say is that most customers want virtualization... the ability to slice and dice the machine to meet the demands of the various workloads. In that realm, Solaris is WAY, WAY behind. Containers are very old technology, that have some benefit, but many, many drawbacks... The hypervisor component of SPARC boxes, for example, using LDOMs and Solaris VM Server for SPARC is about 5 years behind in the management capabilities of IBM/PowerVM and about 10 years behind VMware.... with whom all enterprise vendors are trying to complete.
Simply put, Oracle Enterprise Manager Ops Center is an embarrassing product... it is buggy, inflexible, and downright incapable of providing a stable, reliable and robust virtualization management platform. It is getting better, but my own experience says that there is more work needed.
The Underlying technology for Solaris VM Server and the LDOMs is pretty good... but still not at the the level of IBM LPARs.... heck, just ask this simple question: I want to have fully isolated VMs and share threads between them... how do I do that? Oracle Answer: Containers... but the VMs will share a core OS. *FAIL*. IBM Answer: LPARs can share all CPU resources and if you need containerization, we can do WPARs.
Ultimately, while IBM kicks Oracle's butt in the "real world implementation" of Enterprise UNIX... the "Hottest OS" is Linux... Redhat Enterprise Linux to be precise. More traditional UNIX workloads are being moved to X86 running RHEL than anywhere else.
My $0.02..... from an Oracle and IBM and VMware certified consultant and contractor.
Why would anyone want to share threads between VM's if you don't have to? The reason you share threads on IBM Power is because you don't have enough of them to go around, so the kludge to deal with a hardware resource shortfall is to have the hypervisor time-slice the few threads available. On SPARC, you have enough threads to allocate them fully to VM's, eliminating the time-slicing/context-switching problem. In fact, most SPARC customers assign full cores to VM's, since there are enough cores to go around as well. I really think SPARC has a far superior solution here for enterprise class virtualization.
"I want to have fully isolated VMs and share threads between them"
You only have to look up how many threads you have to play with on current SPARC hardware. You get over a TB of threads on a T5-8 for example. You still want to share threads ?? ( The very first 1RU T1000 back in 2005 had 32 threads ). Trust me if what you're saying was a real problem it would have been fixed a long time ago.
"Containers is old" - Read up on what the linux community is doing with lxc containers now after all these years with bare metal and hypervisor virtualisation. They're coming full circle back to the "old" container technology because there are huge benefits in efficiency. Of course this is not in the best interest of the Virtualization bigwigs so oh no don’t expect this to be news!
Solaris containers have been in production a long time unlike lxc, and the improvements in features and manageability over the years right up to "kernel zones" in Solaris 11.2 is nothing short of impressive ( If you care to look ). It fits a lot of use cases much better than LDOMs, LPARs or any other hypervisor level virtualisation, which all have much higher overheads. But most people rather follow the herd with what everyone else is doing rather than do their own research. Solaris & SPARC haven't been standing still. Most of what you read about Solaris on the internet is old news from experiences on decade old hardware being repeated verbatim by the "experts". Much has changed since Oracle took over in terms of investments and R&D so maybe keep an open mind ?? Wouldn't it be better if there was more competition in the industry ?
My $0.02 .. from an RHCE, AIX and Solaris certified consultant.
"You still want to share threads ??"
Can unused threads from one container automatically be used by another container? If not containers are missing an important aspect of virtualization.
"Trust me if what you're saying was a real problem it would have been fixed a long time ago."
One example of a real problem that has not been fixed a long time ago: Doing something with a container without a reboot. That will be fixed when Solaris 11.2 is out later this year.
"Can unused threads from one container automatically be used by another container? If not containers are missing an important aspect of virtualisation."
Containers have always had the ability to do this. CPU virtualisation in containers has always had the ability to use CPU shares which will allow hardware threads to be shared between various Containers.
LDOMs cannot. But you still miss the point. You're asking for a solution to a problem that does not exist in Solaris LDOMs. Forget about aspects of virtualisation. There's varying designs & implementations. There is no reason to share threads when you have over a TB of threads available. The number of LDOMs on any one such system will be a fraction of this number which negates the need to share threads and thereby add more complexity and overhead to virtualisation.
"One example of a real problem that has not been fixed a long time ago: Doing something with a container without a reboot. That will be fixed when Solaris 11.2 is out later this year."
It has been possible to setup containers with dynamic resource management capabilities for quite some time now. A lot depends on whether it was setup with dynamic configuration in mind or not. It's like setting up an AIX LPAR with dedicated and capped processors and then wanting to change that on the fly. No can do.
Granted that PowerVM has an easier learning curve with a more consistent approach compared to Solaris where there's a plethora of options which can get confusing. An AIX admin can jump from one customer to the other and see an almost standard setup whereas a Solaris admin will most likely have to spend a bit of time figuring out the intricacies of a virtualisation setup. In that respect PowerVM may be better, but feature wise and what you can and cannot do is not really a problem in the Solaris world. In fact you have a lot more options. Much is in the hands of the sysadmin to do things right. AIX and PowerVM is more fool proof and hides a lot of the complexity from the sysadmin.
@MadMike, only because you've pushed my buttons with so many ridiculous statements that I am taking the time to respond. You throw crap into the wind hoping it blows away from you. This time it is coming right back onto you. Solaris is not the hottest OS, as another gentlemen said that would be Linux in general and AIX in the Unix space. Your facts are wrong about Power. A 795 with 256 cores only needs about 64 to outperform a 256 core M9000. Any of the Power7 servers require 1/2 the cores on a staightline race to beat T5, factor in the efficiency of the hypervisor running dozens of VM's and it goes up to 10 - 20X. Just so you don't cherry pick a benchmark and say "look here". You pick your T5 server and I'll bring my largest Power8 server which is 24 cores and let's have at it. Let's run'em to 100% utilization both running Oracle then I'll do another run with DB2 10.5 just to show how much further it scales than Oracle. Did I mention I will be bringing my FlashSystem 840 - don't worry about that though - you will have your 128 core T5 right! By the way, just to show you were wrong, a 64 core 770 supports 1000 VM's (IBM artificially limits it at this number) and can create up to 5120 threads in the server if needed but who would. The point is, you don't know what you are talking about let alone knowing anything about Power technology. By your logic since IBM invented DRAM and HDD, Oracle is "copying" them? How childish! AIX WPAR's are not copied from Solaris Containers. Maybe Containers came from BSD Jails?
With regard to AIX being killed off - what are you smoking? IBM's announcement of Power8 on April 28th supports all 3 of it's core OSes as it has for a decade - IBM i, Linux and AIX. It is as committed to AIX and IBM i as it ever has been. You are hearing about Linux so much because IBM is trying to grow this market. It owns the Unix market with AIX. No matter how much they were to market AIX and PowerVM to some in the open source community they won't budge, so unlike what Sun and now Oracle couldn't do which is embrace open without conditions (ie CDDL), IBM created OpenPower Foundation, is using OpenStack for systems management, added PowerKVM as a option for Linux users in lieu of PowerVM, and added Canonical Ubuntu and Debian Linux to the stable with RedHat and SuSE. With Power8, it runs Java 2X over x86 and Linux (initially Ubuntu) in either Little Endian or Big Endian mode (ummm T5, what say you?) All it takes for LE support is a recompile. (read as period). Ask Ubuntu how effortless it was for them. Take a look at http://benchmarkingblog.wordpress.com/ to see the new #1 benchmarks from these 2 socket servers for SAP, Oracle, Siebel, Java, SPECfp and SPECint. Of course, unlike Oracle's benchmarks where they beat and compare to themselves you may not be that impressed. Last comment. In 2015, Oracle will release a ..... yeah, because they are so reliable in delivering what they say they will. I predict you will see a 2 socket Power8 server with more memory before you will see a SPARC server with 64 TB of RAM. Get back to your basement and clean your room!
"I'll bring my largest Power8 server"
You're comparing an yet to be released Power 8 system with a T5 that's been shipping for at least a year now ? :)
"in either Little Endian or Big Endian mode (ummm T5, what say you?)"
SPARC has been bi-endian since V9 ( since 1993 )
I'll just ignore the rest which is just more BS
You're comparing an yet to be released Power 8 system with a T5 that's been shipping for at least a year now ? :)
Oh he's doing far better than that.
He's comparing a system which comes with a built in 'entitlement' chip whos sole purpose is to do exactly what Oracle is being sued for by Terix. Yep IBM have engineered their competition damaging policy into the next generation of their hardware platform... If Oracle are clever they're just about to receive the opportunity to win back market share in the commercial unix market.
Agree. IBM Power systems and AIX have always required entitlements and licenses to enable anything more than the base OS. A VPD card on the system board stores entitlements for APV ( Advanced Power Virtualisation ) features that has to be enabled. PowerVM (VIO), HA etc all have to be bought separately. Even the xlC compiler costs $$. On top of which you need to buy HMC hardware to manage all of this. Stark contrast to Solaris where everything is included as part of the OS. All forms of virtualisation ( Zones/containers, OVM for SPARC/LDOMS, OVM for x86, Cluster software, OEM/Management software for the entire stack, Oracle Studio/C compilers etc are all free to use and don't cost you extra. There is no concept of locking things down till you pay extra, in the Solaris world. It’s all included with Solaris.
"I doubt that RedHat or IBM give a sh*t about Solaris. It was a great operating system, but it has pretty much stagnated under Oracle and that's shown by the market share it now has."
Agreed, IBM has taken over the Unix market. They don't need any tricks to beat Solaris and have been taking market share from them consistently for a decade. Similar situation with Red Hat, to an even greater extent.
In terms of new developments, patches and fixes is that the overall level of QA and regression testing completed by these development teams is shockingly bad to the point that I now have to consider us as the end client "beta testers" for pretty much everything that comes out - and I do not think that this is by chance either. It now means that I have to add an extra contingency cost to balance that risk on every Oracle ERP and SAP project that kicks off.
Well, if Oracle actually showed some sign of interest in Solaris other than its billability they might have something to gripe about, but if you want a good laugh ask to speak to their resident expert on configuring Oracle RAC on Solaris.
Once he/she's on site, see how long it is before they confess they are really Linux experts with no idea what to do on Solaris.
Oracle doesn't really have a strategy for SPARC or Solaris from where I'm sitting, and would be better off selling that part of the business back off again. All their effort seems to be in Linux-over-Intel these days.
If I didn't know better I'd think Oracle has locked itself in the same battle with Red Hat that Apple did with IBM in the Gaslight Years of Computing.
Oh wait, I *don't* know better.
Yes they would. But nobody wants it. R&D is too costly for the return. That's why Oracle wound up with it in the first place. What they wanted was Java and some other Sun software and they had to take the hardware business as part of the deal.
Maybe, maybe they could pitch it to Lenovo. But at that, I'm not sure they could sell it for $1.
Note that I'm not saying we wouldn't be better off if SPARC could make a run of it. But that ship has already sunk.
>if SPARC could make a run of it. But that ship has already sunk.
Yep happened around the era of the dot com bust. Now when developers have to deal with the SPARC boxes actually found out in industry they are old and slow as hell and a total PITA much more painful to develop on than Solaris on the developers $500 Dell PC desktop.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
