What should Oracle do with Sun
Time for the wisdom of the crowds; what should Oracle do with Sun?
This topic was created by Chris Mellor 1 .
Time for the wisdom of the crowds; what should Oracle do with Sun?
The question should be-what has Oracle done with Sun (since the acquisition)? As far as I am concerned, its converted a failing business model of selling HW at razor thin margins and giving away software for free(Java, Solaris, OpenOffice, etc), which might have drove high revenue (for HW atleast) but drove very little profit (hard to invest in R&D when theres little profit) to a very profitable business, selling only select HW that is viable (SPARC & Eng systems), that’s competitive and that help drives Oracles business model of making profit (hardware and software engineered together) so it can invest in R&D to innovate and continue to compete. Oracle has disinvested in selling x86 servers which have no margin(but still investing in Exa* systems based on x86), selling other companies products like HDS, Fujitsu etc. This clearly has an effect on HW only revenue, but if you look at the profit of the company, profit from HW, its margins and increases in revenue of Oracle, Oracle continues to grow at strong rate as seen by its latest financial reports. The trend is clearly towards converged systems and Oracle is clearly a leader in this trend.
Nobody with any kind of a clue is specking Sun anything anymore.
Talk about squandering a brand ...
Anyone making senseless comments like yours are the only ones without a clue.
I agree with jake, all the sun action now is looking at how to migrate our farms OFF solaris and out of oracle's greedy clutches after their stewardship attempts and price rises so far.
Oracle should... spin off sun and have nothing to do with it, then maybe sun will return to the trusted powerhouse it once was.
We run mostly Linux boxes now, and have been aiming to move off all of our old Sun hardware as time & funds permit because post-Oracle the support has gone down and the price has gone up.
We used to be great fans of Sun, having run various machines of theirs since 1988 and liked the openness and innovative software and hardware, but Oracle are just screwing the customers for all they can get and frankly we don't need SPARC or Solaris, so can and will escape.
Sad really that the winners in business seems to be those hell-bent on screwing their customers, and that customers let them.
I think doesnt matter sums it up well.
Sun/Solaris was great, but they got overtaken on all sides by commodity hardware.
In the late 90s they were charging more for a half-decent videocard (the elite3d and creator3d) then it would cost to spec out a full fledged linux PC with top of the line (at that time, Geforce TI cards) graphics.
That linux box would run circles around the Sun machine, both on number crunching and video performance.
They were (and maybe still are) great in entreprise grade really big iron, but other then that ...
> They were (and maybe still are) great in entreprise grade really big iron, but other then that ...
But that should be a profitable niche, were it not for the fact that some enterprises who used to buy their own kit now just outsource the whole lot - often to oracle.
x86 management is still rubbish in comparison. SPARC haven't been performance leaders for ages, but a consumer-grade x86 chip will also knock the socks off a server-grade x86 chip in terms of performance. Reliability matters in some cases, but as the volume of data grows the less important any one bit of that data is seen to be and it all gets swallowed in generic outsourcing deals at which point the people who really need uptime have lost control of the machine and all that matters are the SLAs.
Most businesspeople are fine, upstanding people. They are NOT the ones who are bribing the cheapest politicians to write the laws they like. It is the least ethical and greediest businessmen who are doing that.
You thought a business was in business to make money? Nice, but I regard that as a naive view. I think it would be better if businesses were in business to stay in business, and of course sustained losses will put you out of business. However, the way the lobbyists have rigged things, all large businesses must grow like mindless cancers just to survive. Moderate success, integrity, any of that old Sun nonsense... Well, you see where that led Sun, eh?
The problem is that there is no such thing as a cancer that is too big to fail. Ultimately the host dies and the cancer dies with it. Cancers are rather mindless that way--but that's how I've always thought of certain businessmen. They just want to die with the most toys.
I'm a 15+ year Sun veteran and now with Oracle and what you are saying makes no sense. So what you are saying is you are jumping out of Oracles greedy clutches where they are making less revenue on HW (meaning they are charging less to customers), have been heavily investing in Solaris innovations that reduce customers costs and moving to Linux where you are in RedHats greedy hands or maybe IBMs greedy hands? Have you seen RedHats rise in revenues? Its from selling services. Wonder why theres such an increase in service revenue for Redhat? Because they are charging more for services and the complexity of Linux is getting worse requiring more support. Why is IBM doing so well? Because more than 50% of its business is services and when complexity is king, service revenue is king. Sounds like you are jumping out of the frying pan and into the flame. Don't get burned.
"So what you are saying is you are jumping out of Oracles greedy clutches where they are making less revenue on HW (meaning they are charging less to customers)"
That is not what it means at all. Revenue = top line sales. Profit = profit. Legacy Sun is making far less in sales, but higher profit margins through price increases and support cuts on the systems they are still selling to people who haven't had time to migrate yet. People were migrating off of Sun anyway. When Oracle decided to increase prices and decrease support, the migrations started to happen at a faster rate.
"Solaris innovations that reduce customers costs and moving to Linux where you are in RedHats greedy hands or maybe IBMs greedy hands?"
Oracle is a Linux first shop themselves. Oracle DB, WebLogic, etc are all developed on Oracle Enterprise Linux and then ported to Solaris and everything else. The flagship Exa systems all run on Oracle Linux (i.e. a fork of Red Hat Linux). Solaris is definitely tier two to Oracle Linux at Oracle.
"Why is IBM doing so well? Because more than 50% of its business is services and when complexity is king, service revenue is king."
Always have to laugh at this one. IBM was/is the ORIGINAL integrated/engineered systems company. They got into integration services because Oracle, Cisco, EMC, etc argued, persuasively, that the integration work was worth the price flexibility and avoiding lock-in. Now that Oracle, Cisco, etc have saturated their component markets, they are making the integration argument. In other words, the same thing they argued against to become as large as they are today. Yes, if you agree to buy everything from one vendor and pay whatever they think is fair, there will be less integration... because you have no choices.
I somewhat agree with your arguments but have several counter arguments as well.
Revenue doesn’t necessarily mean top line sales, it means how much money you are making off of your customers. Revenue certainly doesn’t mean volume (units).
For example, according to the latest IDC WW server sales tracker, Oracle SPARC server *units* out shipped IBM Power servers during CY2011 and SPARC servers out shipped Itanium by over a factor of 2x. Solaris systems out shipped Oracle Linux systems by over 2x with over 4x more in revenue. So over the past 5 years, 52% of the UNIX shipments have been Solaris-based systems and Solaris is the UNIX market shipment leader at 34% for CY11.
And as for IBM "was/is the ORIGINAL integrated/engineered systems company", that may be true in distant past, but since the last 5-10 years have been focused on services, IBM has lost the edge in delivering a truly engineered system based on non-proprietary technologies. Just look at all the gaps in their Puresystems. Which HW technology is "engineered" with which software and does mixing the two make anything faster? IBM's just selling a kit car fully assembled.
"For example, according to the latest IDC WW server sales tracker, Oracle SPARC server *units* out shipped IBM Power servers during CY2011 and SPARC servers out shipped Itanium by over a factor of 2x. Solaris systems out shipped Oracle Linux systems by over 2x with over 4x more in revenue. So over the past 5 years, 52% of the UNIX shipments have been Solaris-based systems and Solaris is the UNIX market shipment leader at 34% for CY11."
None of this speaks to the amount of money customers were being for a server before and after Oracle downfall in Sun revenues. I am sure Sun did ship out more Sparc servers than Power or Itanium. IBM Power and, to a lesser extent, Itanium ship fewer, larger servers. Getting more smaller, less powerful servers for a lower price doesn't matter as you cannot run the same amount or types of workloads on that increased number of smaller severs. IBM and, to a lesser extent, Itanium, are primarily massive DB servers being compared to Sun's web servers. Regardless, the unit shipment count does not address the price paid for systems and support to the end Sun user, which is higher. Losing market share does not make the costs to the end user go down in any space. Those are just fewer customers to spread the fixed costs of plant/equip, design, back-office operations, etc across. As Oracle has a smaller market but has increased Sun's profitability, there are only two ways to do that: increase the prices to the customers or cut operations.
"IBM has lost the edge in delivering a truly engineered system based on non-proprietary technologies. Just look at all the gaps in their Puresystems."
IBM does have awesome "engineered" systems. System z is a private cloud before anyone came up with the concept of a private cloud that is tightly integrated with DB and the stack. Netezza, although IBM acquired and did not create it, will smoke Exa on pure OLAP workloads. The difficulty for IBM, and even more so for HP and everyone else, is that Oracle has a dominant position in the very important DB market. As such, they are able to use their DB control to "engineer" hardware and utilities which are tightly coupled with their DB software. Oracle wants everything to be about their ubiquitous DB. If that is the case, I wish Oracle would just admit what they are clearly doing, building an Oracle solution from hardware through applications (the "IBM of the 1960s" as Larry says). General purpose Sparc - Solaris and the other Sun brands do not help them with that vision which is why they are being cast aside. The irony of the whole thing, as I mentioned, is that it was Oracle, Cisco and others who originally were the strongest advocates for integrated as opposed to engineered systems because, at the time, it was in their interest.
What gaps? PureSystems is not going to be as integrated as Exadata, but that is because it is not a lock-in from tip to toe. IBM has kept the option open of using whatever hypervisor, OS, back-end storage, application stack you like. It is the best option going for people who want to preserve some choice and not cast their entire lot with one vendor.
"Which HW technology is "engineered" with which software and does mixing the two make anything faster?"
Not arguing that building the software and the hardware as one appliance will not improve performance and lower administrative costs. I am arguing that when you go all-in with one provider, with no choices, there is not much incentive for that provider to continue to offer competitive pricing after you are all in.
You can put lipstick on a pig dog, but it's still...
Hold on a minute, what?
Erm, why is this comment page seemingly detached from the story that points to it at http://www.theregister.co.uk/2012/06/20/oracle_sun_revenues/ ?
Your basis for calculation splits hardware revenues into two categories. What do IBM's, HP's or any other vendors figures look like if you apply the same split to hardware sales in competition with Sunacle?
"What do IBM's, HP's or any other vendors figures look like if you apply the same split to hardware sales in competition with Sunacle?"
IBM = Looks really good
HP = Looks really bad
I agree. This whole analysis is fluff. The bottom line is that Oracle continues to be profitable and continues to drive profit and revenues on both HW and SW. Selling other peoples products & technologies doesn’t fit into this equation and trying to figure out how much revenue (and if there any loses compared to pre-Sun days) has been made on HW is irrelevant. Why hasn’t there been any analysis on IBM's massive mainframe revenue declines or HP's Itanium declines vs PA-RISC days? Its clear it’s a new world with new trends, including the cloud and convergence and clearly Oracle has turned around the Sun HW part of the business and is remaining successful with continued growth after the acquisition.
Isnt this bit of a spurious comparison.
Sun's hardware business was dying on its feet BEFORE oracle.
Its entirely likely that its future as a standalone was in Engineered systems away. It seems it bit silly to exclude these since the guts of the hardware is ex-Sun stuff.
So really the headline is Oracle's Hardware Business in transition from commoditised servers to value added engineered systems.
Change Oracle to Sun in that last sentence and would it look as bad?
"So really the headline is Oracle's Hardware Business in transition from commoditised servers to value added engineered systems."
Yes, if you consider Sparc - Solaris a commodity server.
The engineered systems are _MOSTLY_ not SPARC based, but rather x86 servers running Solaris x86 and Oracle application workloads... They did mention the SPARC Supercluster, though, which is (surprise, surprise) SPARC based... :)
Oracle's excuse for the massive fall-off is that they are just ditching the low margin x86 servers (not including Exa, obviously) that Sun used to just about give away. That couldn't possibly be the case. As listed below, Sun's x86 server sales at their peak were only $707 million per year. Even if Oracle stopped selling any general purpose x86 servers, it would not come close to explaining the gap. It is Sun Sparc, ZFS arrays, StorageTek, and basically everything non-Exa that has fallen off. Oracle needs to stop messing around with Sparc. Everyone hates Sparc, even the Solaris die hards. They should just port Solaris to x86 and kill Sparc. Oracle is just making themselves look ridiculous in comparing Sparc to Power. No one believes those absurd benchmarks who has worked with both Power and T Sparc, which then causes them to question everything Oracle says about Sun.
Anonymous Coward: "They should just port Solaris to x86 and kill Sparc."
What are you talking about??? Solaris has been ported to x86 for a very long long time! More than a decade ago , in fact I think already in the previous millenium.
In fact, many companies run Solaris on x86/64 server and many run it on vmware even!
What Sun refused to do was port Solaris to ITANIUM ... but even Redhat agrees with them on that decision now.
With the exception of Sparc Supercluster, Oracle's engineered systems have nothing to do with Sun tech. Exadata, Exalogic, etc run on Intel x86 processors with Oracle Linux, not Sparc - Solaris. As Sun was not selling these systems prior to the Oracle acquisition (Exadata was HP based), it doesn't make sense to include systems they were building without Sun as part of legacy Sun. If Oracle just wanted to build Exa systems, there was no need to acquire Sun.
We laugh at Oracle when they try to sell us any type of hardware. I was out for drinks with an Oracle rep and he was complaining how when they do sell hardware it gets bundled with the software. The software has to be discounted behind the scenes so that hardware gets sold at zero discount. The numbers you are seeing from oracle is inflated by playing with discounts and bundling show what is rightfully software revenue as hardware. Those hardware numbers should in reality be cut in half again.
All the technology (database/weblogic) sales reps and applications reps hate the hardware reps and cut them off and try to not allow them into their accounts.
Agree. From the interactions I have had, Oracle software reps want nothing to do with the Sun hardware reps. They try not to let them into their accounts because it will involve some software discounting to make the hardware, which they do not care about, look attractive. Big time cultural conflict and conflict of interests. If the Oracle software reps were in charge, everyone would run on Itanium. Not because they think Itanium is any good, but because they know it isn't. If someone wants to pay a fortune in Oracle licensing to run on a ton of core factors of HP-UX Itanium, don't get in their way as far as the software reps are concerned.
All the Oracle software sales reps I have encountered for years years been saying low socket count bare metal Wintel boxes, the cheaper the better. And then smack RAC ontop of that.
The reason is that they then get a bigger part of the IT budget pie.
I suppose that would be another angle, selling the RAC licenses. Unix systems do cut into Oracle RAC sales. I was being sarcastic with Itanium, but my point is that Oracle software reps only care about hardware if it allows them to sell more licenses, either on a straight core factor or with additionals like RAC. The reason they don't like Sun is that Oracle wants to make HW look as good as possible at the expense of software. It is all behind the scenes, but if someone buys a bunch of Oracle HW and software they will count the HW being sold at full list and full margin and the software as being heavily discounted... when the SW reps know that the customer usually only bought the HW because it was wrapped into the SW they were going to buy anyway.
Well when you deal with IT vendors that supply software and hardware, they are always trying to get an as big as possible slice of your IT budget pie.
Hence if you as a software company can convince your customer to use low grade cheaper hardware, then there is more IT budget left to be spend on software. Hence you have a chance to get more revenue.
Personally I'd rather spend money on quality hardware than on additional software licenses that aren't really needed. Again the hardware will bring value, added software licenses won't.
Sun has been dying for more than a decade that I know of, it is the last of the 'big iron' firms.
IBM re-invented itself about twenty years ago very successfully but Sun failed to really embrace change.
IBM is the last of the big iron firms. As with mainframe and minis before it, IBM will come to be the last player in a smaller Unix market.
Not quite - IBM is now a collection of companies e.g. Rational Sofware, printers, Storage Subsystems Division (SSD), PowerPC etc.. Sun is still an old monolith.
Yes its true. Sun died a long death after the dot com era because they couldn’t figure out how to make money off of giving away software for free (Solaris, Java, Openoffice and many others) and the hardware business that they were driving to, was becoming more and more commoditized -it came to a point of trying to sell hardware without much differentiation (especially x86) and making razor thin margins which made it difficult to remain profitable and invest in R&D. IBM has reinvented itself by changing from a HW focused company to a services focused company selling hardware and software to drive services. Redhat, HP and many others are focused on the same business model. And as long as IT complexity remains king, services revenue will continue to grow. Oracle is headed in a different direction of trying to eliminate the need for many of those services by engineering the entire stack and selling you a Red box that’s easy to manage and maintain over the lifecycle, reducing management and administration costs a long the way.
The services vs. integrated argument doesn't hold. IBM only became a large services company because of Oracle and former friends (Cisco, Sun, EMC, etc) arguing that distributed stack with flexibility and choice was the best model. IBM was the original company who argued that getting away from consolidated, engineered systems, i.e. the mainframe, would create tremendous amounts of additional admin and integration work. Now that Oracle has basically saturated the database market, they are using the same argument IBM used as though they discovered something new. It is really on another level of absurdity to argue that IBM doesn't like large, integrated consolidated systems as a model. They were the ones who invented it.
I've placed no less than three calls to Oracle to set up maintenance contracts for my Solaris boxes. That's the only way you can download Recommended clusters and other updates. I've yet to have one call returned.
I'm running Solaris on commodity hardware, and it runs damned well. I've always preferred Solaris over Linux because it HAD a reliable commercial backing and I just like it better. It's been a rock-solid operating environment on any hardware, be it Sun SPARC or x86 or commodity. But it especially runs well on Sun hardware. I've been given anecdotes about a local college switching from Sun hardware, which took massive beatings during student registration and financial aid season, to Linux machines. The latter went down spectacularly.
In any case, I was concerned Oracle would mis-manage or just kill Solaris out-right when it bought Sun. I was particular disappointed that Sun couldn't adapt to the New World. Remember when it was the "dot" in "dot-com?"
Paris, goes down spectacularly.
Sounds like you're not calling the right folks at Oracle-selling maintenance is quite high on Oracles priority list just like it is for IBM, HP, Redhat and others where profits are high.
As for the new world, see my post at end of this comment section
I work with Sunacle engineers. The impression they get from the top is that anything non Exa- is a waste. The Service and support on non Exa- kit is so poor that the decision makers don't want to buy any Sunacle engineered systems because their current impression of the business is of poort service and support.
I haven't met anyone interested in buying a new Sun server/storage in years. Are there really that many legacy Sun accounts out there that are refreshing to create $5 billion per year in sales?
Oh right, well if you've not met them, they obviously don't exist.
I know a lot of people who are buying AIX. I know more who are buying x64 with RHEL. I know a few who are looking at Exadata.... which, I suppose, is also x64 and Oracle RHEL. I even know some who are still plugging away with HP-UX. I don't know a single person who is buying new Sparc servers.
I know people who have them, but they are all in the process of moving away and not refreshing anything. They must exist because Oracle is selling that gear to someone.
Not scientific, I didn't run a survey, I am just saying it is strange that they are still selling $2 plus billion worth of new gear a year. What is the name of a large Sparc shop that is still refreshing and adding?
If you really want to see the impact of Oracle on legacy Sun, look at 2008, the last full year for Sun prior to the announcement of the Oracle acquisition. I know the acquisition did not get approved until 2010, but it was announced in 2008. Sun had systems revenues of $8.6 and support revenues of $4.0 or $12.6 for systems related revenues in 2008. Software and other stuff may fall into support revenue, but we can compare systems sales without support pretty accurately from then to now. Oracle just did $900 million in systems revenue last quarter, as compared to the $2 billion plus they were doing per quarter in 2008. As the article states, you need to subtract Exa to get a true picture. If Exa = $1.2 billion per year, that means that legacy Sun is somewhere in the $600 million per quarter range. Subtract the $100 million per year for Pillar and it comes to $2.3 billion. In other words, in a little over three years legacy Sun systems have gone from $8.6 billion to about $2.3 billion per year in new system sales.
For those who are Oracle hard liners and accept that this was just dropping x86, Sun's x86 server revenue in 2008 was $707 million.
They should be more pragmatic and devote more resources to the M-Series(Sparc64) line.
Never going to happen. It is Fujitsu resell and Oracle doesn't resell. They are moving off of Sparc64 and on to T Sparc for the M-Series.
Don't know what the fuss is about. Oracle bought Sun, because they wanted Sun's IP, customers and business - not Sun's hardware, which they couldn't give a hoot about. Why is it a shock that Sun HW sales are going down? Even when it had proper investment, it was already too slow compared with x86 for the low-to-mid segment.
No shit, Sherlock..
And unlike HP Oracle is smart enough not to continue to dump cash in a dying architecture nobody is buying. Sadly still two years from now there will still be more SPARC chips being sold than Itaniums (together they will sell as as many in a quarter as ARM sells in China in 17 seconds or something lol).
Sadly? And there will be Itaniums being sold two years from now??
From customers I've heard that Oracle is pushing SPARCS to run their Enterprise Database.
Although it cant match Power6+ or latest X86_64 on single thread speed, the Sparc architecture still offers great benefits when it comes to hardware redundancy, storage management and a rock solid OS in the bottom with some nice features you will never see on RHEL (DTrace, ZFS, Fault manager, Service Manager to name a few).
They discount the oracle core licenses so they become competetively regarding price/performance.
The hardware price is nothing compared to the $47.500 per processor list price, and AFAIK no other processor beats the 64threads on T4 or 128threads on T3.
Add up the discounts and the most attactive system to run a large oracle database is on Sparc/Solaris.
The next SPARC T5 will have 128cores and 16K threads and IO backbone to sustain it.
Singlethread will be mediocre but no one else
I work with databases that grows with >10mill rows each day, and I've never seen any need for faster cpus, it is mostly IO and memory. If a single thread is slow, just run them in parallel and you get linear speedup.
When you run queries you need lots of indexes to back you up which makes you IO bound.