No, you just dismiss the failures because of the successes, as if being successful at servers would be a bad thing.
They had a product in XServe, but never really realized it's potential. Shitty SLAs, blah hardware choices, and expensive. More of a hobby. I'm not saying they throw darts at a board, I'm saying they had a product that made people think they would evolve into the enterprise market, and failed at it.
You are saying "who cares", which I get, I just don't understand how you can say it in this thread. The thread is about the death of Xserves, Apple's only mediocre attempt at rack servers. Steve obviously knows how shitty of a job they were doing with them, and that's why they're dead.
No need to be a douche about it....I'm guessing your island is Manhattan?
So they had what by your lights was a fairly horrible product but it's still a bitter pill to swallow that it's been discontinued because "people" thought it would "evolve"?
It's not just you, it's a lot of posts in these various Xserve threads, that seem to be saying that the Xserve was never that great, had ludicrously underbaked support, and didn't do anything that can't be done much better by commodity hardware, but that killing it is some kind of insult to Apple's users and a grim harbinger of a trivial future, all because they were "supposed" to make it better, at some point.
Really not following that. There's a lot more to a successful server business than decent hardware and 24/7 support; it's probable that after evaluating their market position Apple decided that they should fish or cut bait, rather than hang around half-assedly.
So it seems to me that the real conversation should be about whether or not it would be good or reasonable or advantageous for Apple to be in that business at all, and not lamenting the loss of the (apparently) grossly underperforming Xserve.
Of course, "Should Apple be making a big push into the enterprise server space" is more difficult to frame in near apocalyptic terms, so maybe that's it.
The XServe didn't sell, that is why it is being dropped.
As to cloud computing, in many cases the concept is grossly over sold. Few companies would want to entrust their computing needs entirely to a third party. Plus all this cloud excitement doesn't take into account real issues like latenecy and bandwidth issues.
Quote:
Originally Posted by jkichline
Does anyone else think that the discontinuation of the XServe product is a harbinger of Apple's push into cloud computing? While a Mac Mini or a Mac Pro is great for typical office networking needs, it isn't a great solution for large-scale hosting. But if Apple is moving in that direction with it's new data center, perhaps that should be less of a concern. Thoughts?
I'd say you are obsessed with the idea of cloud computing! Honestly it isn't all that it is cracked up to be an exposes companies to additional problems they wouldn't have other wise. You need to understand that the whole concept of cloud computing is being pushed by marketiers trying to make money. By definition they don't have your best interests at heart.
I'm not trying to say Apple won't have a cloud play just that it is wise to realize that it is likely not in your best interest as a user. Especially if that play involves streaming of music or movies to your "I" device. It is nothing more than a plan to get you to pay for services instead of simply selling you an "I" device with suitable storage in the first place. Cloud services are a way for Apple to tap into your wallet 24/7. Part of the reason you don't see a mad rush of business to such services is that they need to evaluate the move based on economic principals. Consummers however are in most cases stupid and get involved in scams without thinking.
Business will look at external services when it makes sense. Google mail is one example of a successful out sourced service. Companies though don't just jump into these things as the change has to be cost effective and meet the other needs of the company. Those other needs can be significant.
This isn't clear for me... Why on the hardware that is intended to work as a server there is a video card with 1Gb of RAM ???? Looks like the new server standard for installing OS
If enterprise is out think about where else these servers might be going. Being able to support OpenCL is a smart move for many server markets. In my estimation one of Apples big mistakes with its XServes was not going after the scientific computing market. Admittedly a smaller niche than corporate data centers but one that plays to Apples strengths. Interest in XServes could have been far greater if they had integrated an OpenCL compatible GPU into the motherboard when OpenCL debuted.
We can only hope that Apple has something else up their sleeves for file serving though. The Mini is a bit of a joke, mainly because of servicing issues and ultimate capacity. On the flip side the Mac Pro is way to expensive for simple file sharing duties.
In any event I think this move will hurt Apple in the long run. Unless they can come up with a compelling alternative. It can be likened to going to a hardware store to buy some spikes and finding nothing but common nails in the fastener section. Soon you find another hardware store to buy your spikes and then you start to shop there more because the other place can't supply you with the odd spike when you need it.
That is of course one way to look at it. The other is that in many cases the corporate data center and the desktop are often served by hardware from different manufactures anyways. So this does not in anyway imply that Apple isn't interested in the enterprise. Rather it indicates that they aren't interested in competing with IBM mainframes, blade servers or Linux clusters. Since they don't have a credible platform why bother.
Too many people are acting like the Xserve is unsuccessful because they do not see it in businesses all over the place. The fact is that the Xserve is most popular in education, K12 and higher education for management (open directory) as well as for Xsan/Final Cut Server installations. The Xserve is a very nice 1U machine, and I will honestly be sad to see it go. It is a reliable product that is built very well and did what it needed to do. Mac OS X server also offers great web server (wiki/blog), podcast producer, and more. You are all discussing this like this is the death of Apple in enterprise, but Apple does not kill a product line without some strategy in place. This is not because they did not sell the product, as all they would have to do is make the server a great tool for iOS management and deployment and they would sell thousands. Apple will not continue to make a product if they do not feel they are innovating in the space and they do not feel it is part of their overall strategy. A perfect example are the RAID storage systems. Apple used to make APple branded system, but they then stopped and had Promise make all of the systems since they felt other companies could do a better job.
There is a strategy here, it is just that none of us can see it clearly yet.
So what's the future of OS X Server? Especially with no true servers to run it on?
Seems to me Apple should update OS X Server to run on ANY server platform.
It will be interesting to see if Apple continues OS X Server development for use with smaller workgroups where a Mac Mini or a Mac Pro might be viable hardware mated with the GUI admin tools. It is hard to see what benefits OS X Server brings to large-scale deployments like data centers where the admins prefer to live in the CLI anyways.
I can't really see Apple licensing the server software to other H/W manufacturers (like HP or Dell) but if Apple wants to make OS X Server available for large-scale deployments I can see them licensing to system integrators like Unisys who would then be responsible for implementation and support. A company like Unisys or IBM is much better structured to provide the support essential for these types of large deployments. I would not be surprised if this is part of the deal struck between Apple and Unisys.
except that mac mini's have no management capabilities and no RAID or hot swappable drives
All of which add to the complexity of the server. Besides you mis the point of a Mini, you hot swap the whole computer. In many cases the swap is many times faster than servicing a 1U server.
Quote:
with our HP servers i have console access via my iphone. i can even remotely press the power button on a server via my iphone and mobile safari. they don't even need KVM's
Nice security there!
Quote:
a 1U HP server can take up to 8 hard drives and 192GB of RAM. you'll need a whole rack full of mac minis for that.
Honestly how many web sites need that sort of big iron? Even a Mini can support multiple sites. For that matter a Mini with a little 64GB SSD can support many networking needs.
I'm not here to discount the need for big iron in many cases, rather you seem to discount or completely dismiss the idea that the majority of Internet server needs are rather simple. Currently a dual core Mini can sleep at around 7 watts yet provide fairly robust perfomance when needed to. The fans on most servers take more energy than the Mini.
Quote:
by the time you pay more money for power, KVM switches, data center space, more racks, you have wasted a lot of money compared to buying Dell/HP/IBM servers
Nope! You want to believe the big iron systems of the past are the solutions for the future, in many cases they are not. The reason being is that big iron systems are not economical it many different senses. Just look at the space demand of a typical 1U server which are generally huge foot print wise. Your arguements about power are just silly.
Like it or not the future of the server business is in very small and low power devices. Expect to see Minis and i86 hardware replaced by ARM based servers in the near future. It is all about being green and competitive. The data ceneter that can move the most data at the lowest cost wins. Space and electrical power these days are bigger issues than raw server power.
Think about servers built around ARM based cards, maybe 3x8 inches that have your network interface, CPU, flash storage and power supply all integrated on that board. Your entire server might hit 5 watts full load. If you don't like ARM look at AMDs new Bobcat based processors. In essence a mini blade server. Mac Minis are currently a approximation of this approach.
Quote:
and $1500 will buy me a 5 year 24x7 4 hour warranty from HP. with a mac mini i have to take it to an apple store for simple things like replacing hard drives
Anybody with any sense would be replacing those drives themselves. Honestly you can buy a new Mini plus several replacement drives for that sort of money. Further if you are spending that sort of money on service contracts for every server you have you are wasting a lot of cash.
Too many people are acting like the Xserve is unsuccessful because they do not see it in businesses all over the place. The fact is that the Xserve is most popular in education, K12 and higher education for management (open directory) as well as for Xsan/Final Cut Server installations. The Xserve is a very nice 1U machine, and I will honestly be sad to see it go. It is a reliable product that is built very well and did what it needed to do
So?
Seriously so? If the money isn't there how long should Apple continue to support the device?
Quote:
. Mac OS X server also offers great web server (wiki/blog), podcast producer, and more. You are all discussing this like this is the death of Apple in enterprise, but Apple does not kill a product line without some strategy in place.
Well I will give you this much, this will have no effect at all with respect to Appple in the enterprise. However they killed this product due to the lack of demand.
Quote:
This is not because they did not sell the product, as all they would have to do is make the server a great tool for iOS management and deployment and they would sell thousands.
not a remote chance of that happening
Quote:
Apple will not continue to make a product if they do not feel they are innovating in the space and they do not feel it is part of their overall strategy. A perfect example are the RAID storage systems.
Nope both XServe and XRaid where innovative (and expensive). Both died due to lack of sales and a lack of interest from Apple.
Quote:
Apple used to make APple branded system, but they then stopped and had Promise make all of the systems since they felt other companies could do a better job.
There is a strategy here, it is just that none of us can see it clearly yet.
It is called dumping the line that is not performing.
Seriously so? If the money isn't there how long should Apple continue to support the device?
....
Nope both XServe and XRaid where innovative (and expensive). Both died due to lack of sales and a lack of interest from Apple.
It is called dumping the line that is not performing.
Too many people think that Apple is run by a "mercurial CEO" who makes one arbitrary decision after another rather than seeing them as the self-disciplined organization they have become over the past decade.
VMWare and hardware in general has grown a lot in the last 3 years. i bet this is all being transitioned to one vmware cluster if it hasn't been done already
VMWare isn't all it is cracked up to be. Besides even if you accept VMWare you can run it on a Mini too. Lets face it some sites are so low in traffic that you could host several on a Mini through virtualization.
The use of Minis isn't an attempt to have one site per unit from what I can see. Rather it seems to be a credible attempt to get around the issues involved in running big iron servers. Google may be able to afford to design optimized hardware but companies like this can't. A Mini though gives them many of the same advantages in that the platform is low energy usage, effectively compact yet surprisingly powerful for its size.
Take a good look at a Mini motherboard. In many ways it is relatively free of extraneous hardware much like a server motherboard.
All of which add to the complexity of the server. Besides you mis the point of a Mini, you hot swap the whole computer. In many cases the swap is many times faster than servicing a 1U server.
Nice security there!
Honestly how many web sites need that sort of big iron? Even a Mini can support multiple sites. For that matter a Mini with a little 64GB SSD can support many networking needs.
I'm not here to discount the need for big iron in many cases, rather you seem to discount or completely dismiss the idea that the majority of Internet server needs are rather simple. Currently a dual core Mini can sleep at around 7 watts yet provide fairly robust perfomance when needed to. The fans on most servers take more energy than the Mini.
Nope! You want to believe the big iron systems of the past are the solutions for the future, in many cases they are not. The reason being is that big iron systems are not economical it many different senses. Just look at the space demand of a typical 1U server which are generally huge foot print wise. Your arguements about power are just silly.
Like it or not the future of the server business is in very small and low power devices. Expect to see Minis and i86 hardware replaced by ARM based servers in the near future. It is all about being green and competitive. The data ceneter that can move the most data at the lowest cost wins. Space and electrical power these days are bigger issues than raw server power.
Think about servers built around ARM based cards, maybe 3x8 inches that have your network interface, CPU, flash storage and power supply all integrated on that board. Your entire server might hit 5 watts full load. If you don't like ARM look at AMDs new Bobcat based processors. In essence a mini blade server. Mac Minis are currently a approximation of this approach.
Anybody with any sense would be replacing those drives themselves. Honestly you can buy a new Mini plus several replacement drives for that sort of money. Further if you are spending that sort of money on service contracts for every server you have you are wasting a lot of cash.
You could add, 192GB of RAM is much better served in a Workstation with 2 or 3 GPGPUs running Engineering, Physics, Biology, Network simulations, etc.
You bad mouth Jobs and then you give us the crap below as a justification. Sadly it looks like you have more issues than Steve ever had.
Quote:
Originally Posted by theolein
For those of us who have been running Apple's XServe for years, this news, while not entirely unexpected after Apple discontinuing the XRAID in 2008, is pretty damn bad.
So you suspected for 3 years this might happen and now you are angery!
Quote:
Mac Pros and (lol) Mac Minis are not an option in the data center.
Niether was the XServe for most businesses.
Quote:
We've been having an emergency meeting about this
You have to be kidding you had an emergency meeting over this. That has to be the biggest joke that I've seen posted in these forums in years. Seriously the discontinued a model and gave you a whole quarters warning, this is not something that requires an emergency meeting. Not unless you make a habit of building moutains out of mole hills.
Quote:
and we've decided that playing catch up with Apple's lack of stability and Steve Jobs' mood swings all the time (Classic->OSX, PPC->Intel, Software, software, software and now this) is just not worth the cost any more
Are you ignorant or just trying to pull our legs. The Classic to OS/X transition had to happen because everybody was rejecting the old and brittle OS. Like wise with the transition to i86 which had to happen because of the lackluster development in PPC land. As to software what is wrong there, Apple has delivered a state of the art GUI to build any piece of software you could want. Further third party development is going very strong right now.
Quote:
and we'll be transitioning to Windows in the next two years, on the client side as well.
I need to find out who you are so that I can write the board to detail your rash and unprofessional behavior. Seriously this type of thing has sunk companies. There is no bit of sound reasoning or justification for making a decision like this so fast.
Quote:
I'm pretty sad about this,
You sad - BS. It appears that you or someone at your place of work was just looking for a reason to get rid of all Apple hardware.
Quote:
but the plus side is that we won't have to deal with Apple's (lack of) support anymore, which sort of makes it less painful.
OK Be aware that you are making a big mistake. Microsoft is just about the worst possible choice you could make.
You see what is pathetic here is that many people claim that one issue with Apple in the corporate world is that they don't communicate. Yet the minute they try to do the right thing and pre announce a change in the product line up they get crapped on by organizations like yours. You can't have it both ways.
In any event lots of luck with that Microsift based system! The grass isn't any greener on the other side of the fence. Plus we could have a very very long talk about failed Microsoft software initiatives. In the end I suspect you will find out that your very rash behavior will result in far more problems than simply addressing the server issue in a mature fashion.
Yeah! The cowards! How dare they pick and choose what markets they compete in!
I have no problem with Apple choosing what markets to compete in. I do have a problem as a strong supporter of enterprise Apple, being f**ked over with close to no notice. I also am concerned that a company with billions in the bank can't seem to figure out how to address a potentially lucrative market. Apple's 'enterprise push' seems pretty weak without a server product. What they're actually doing is selling gadgets to large enterprise, and semi-servers to SMBs.
. Apple's 'enterprise push' seems pretty weak without a server product. .
There is no real enterprise push for OSX. There is no reason to use macs and osx in the enterprise and it is just a dream of fanboys (in my opinion). In most cases it would be a stupid decision to use macs in enterprise and probably only fanboys would do it. With Apple you have no planning reliability. There are more disadvantages (expensive, bad support, bad backward compatibility, short security support, glossy, ... ). In the enterprise you are working with programs and not with good looking computers. Operating System? Just a means to an end. And in that case Windows e.g. is the better solution: Windows Server with active directory, Exchange, Share Point Server, Office, Windows, ... with Citrix, VMware and so on...
It is all about iOS, when Apple is talking about the enterprise.
The lights-out management, the dual power supplies perhaps? Let me guess, you don't know what lights-out management is, do you?
Oh man. RAID 0 and RAID 1 are for end users, not for data centers. Most data centers run with hybrid RAID 50 or 60 etc. We're not talking about your home office here.
I don't blame Apple for leaving a market they were competing badly in (mostly for lack of trying, IMO)). It would, however, have been nice if Apple had been a little bit more consistent.
You don't know how dumb you make yourself look with comments like that.
Time Machine is an end user/home user backup tool. It's not something that you would use in a data center, where the requirements are at another level entirely.
theolein, I feel your pain. Very few folks here understand the datacenter environment.
Too many people think that Apple is run by a "mercurial CEO" who makes one arbitrary decision after another rather than seeing them as the self-disciplined organization they have become over the past decade.
I just responded to one guy that seems to have this attitude. He claims an emergency meeting about the drpped XServes resulted in a decision to drop Apple entirely both server wise and client workstations. This is complete garbage as you can't make a business decision like this in one day on limited information.
The proble is he goes on to accuse Apple of doing the same thing repeatedly. As an example he gives the Classic to OS/X transition which is absolute garbage. Classic was killing Apple as a computer maker and everyone in the industry should know that. Further Jobs was able to come in, take control of the whole operating systems mess at Apple and quickly get them on track with OS/X. So over the course of the years developers have had only two GUI toolkits to target. Yet this guy wants to switch his company over to MicroSoft software, where I don't even know anymore how many GUI toolkits there are to target.
I'm actually surprised at the indications in this thread of rash and emotional people in business. Really these people should be flipping burgers someplace. Even scarier is that some of these companies might be publically traded. One thing that is expected out of public companies is fisical responsibility and frankly I don't see how anybody can make a responsible decision in less than 24 hours based on this news.
Oh by the way I really don't want XServe to go away but I do expect Apple to make money. If XServe doesn't do that for them then to bad. When it comes right down to it companies will be replacing these boxes with hardware that costs half as much and probably oerforms 2x better. If they are smart they will, but frankly pricing of server hardware from some companies is a little stiff.
Frankly It looks like half the people in this thread feel so damn special that they think they have the right to demand whatever sort of hardware they could want from Apple. Makes you wonder if they are all democrates.
No. In most cases MS is the BEST choice in the Enterprise.
MS >>>>>>>>>>>>>>>> Apple
Edit:
Job title
Windows Systems Engineer
The reasons are clear your whole system can be easily compromized by viruses, bugs and faulty hot patches. That and the fact the MS seems to have lost its rudder with respect to many recent software initiatives, Silverlight being the most recent.
Comments
No, you just dismiss the failures because of the successes, as if being successful at servers would be a bad thing.
They had a product in XServe, but never really realized it's potential. Shitty SLAs, blah hardware choices, and expensive. More of a hobby. I'm not saying they throw darts at a board, I'm saying they had a product that made people think they would evolve into the enterprise market, and failed at it.
You are saying "who cares", which I get, I just don't understand how you can say it in this thread. The thread is about the death of Xserves, Apple's only mediocre attempt at rack servers. Steve obviously knows how shitty of a job they were doing with them, and that's why they're dead.
No need to be a douche about it....I'm guessing your island is Manhattan?
So they had what by your lights was a fairly horrible product but it's still a bitter pill to swallow that it's been discontinued because "people" thought it would "evolve"?
It's not just you, it's a lot of posts in these various Xserve threads, that seem to be saying that the Xserve was never that great, had ludicrously underbaked support, and didn't do anything that can't be done much better by commodity hardware, but that killing it is some kind of insult to Apple's users and a grim harbinger of a trivial future, all because they were "supposed" to make it better, at some point.
Really not following that. There's a lot more to a successful server business than decent hardware and 24/7 support; it's probable that after evaluating their market position Apple decided that they should fish or cut bait, rather than hang around half-assedly.
So it seems to me that the real conversation should be about whether or not it would be good or reasonable or advantageous for Apple to be in that business at all, and not lamenting the loss of the (apparently) grossly underperforming Xserve.
Of course, "Should Apple be making a big push into the enterprise server space" is more difficult to frame in near apocalyptic terms, so maybe that's it.
As to cloud computing, in many cases the concept is grossly over sold. Few companies would want to entrust their computing needs entirely to a third party. Plus all this cloud excitement doesn't take into account real issues like latenecy and bandwidth issues.
Does anyone else think that the discontinuation of the XServe product is a harbinger of Apple's push into cloud computing? While a Mac Mini or a Mac Pro is great for typical office networking needs, it isn't a great solution for large-scale hosting. But if Apple is moving in that direction with it's new data center, perhaps that should be less of a concern. Thoughts?
I'd say you are obsessed with the idea of cloud computing! Honestly it isn't all that it is cracked up to be an exposes companies to additional problems they wouldn't have other wise. You need to understand that the whole concept of cloud computing is being pushed by marketiers trying to make money. By definition they don't have your best interests at heart.
I'm not trying to say Apple won't have a cloud play just that it is wise to realize that it is likely not in your best interest as a user. Especially if that play involves streaming of music or movies to your "I" device. It is nothing more than a plan to get you to pay for services instead of simply selling you an "I" device with suitable storage in the first place. Cloud services are a way for Apple to tap into your wallet 24/7. Part of the reason you don't see a mad rush of business to such services is that they need to evaluate the move based on economic principals. Consummers however are in most cases stupid and get involved in scams without thinking.
Business will look at external services when it makes sense. Google mail is one example of a successful out sourced service. Companies though don't just jump into these things as the change has to be cost effective and meet the other needs of the company. Those other needs can be significant.
This isn't clear for me... Why on the hardware that is intended to work as a server there is a video card with 1Gb of RAM ???? Looks like the new server standard for installing OS
If enterprise is out think about where else these servers might be going. Being able to support OpenCL is a smart move for many server markets. In my estimation one of Apples big mistakes with its XServes was not going after the scientific computing market. Admittedly a smaller niche than corporate data centers but one that plays to Apples strengths. Interest in XServes could have been far greater if they had integrated an OpenCL compatible GPU into the motherboard when OpenCL debuted.
We can only hope that Apple has something else up their sleeves for file serving though. The Mini is a bit of a joke, mainly because of servicing issues and ultimate capacity. On the flip side the Mac Pro is way to expensive for simple file sharing duties.
In any event I think this move will hurt Apple in the long run. Unless they can come up with a compelling alternative. It can be likened to going to a hardware store to buy some spikes and finding nothing but common nails in the fastener section. Soon you find another hardware store to buy your spikes and then you start to shop there more because the other place can't supply you with the odd spike when you need it.
That is of course one way to look at it. The other is that in many cases the corporate data center and the desktop are often served by hardware from different manufactures anyways. So this does not in anyway imply that Apple isn't interested in the enterprise. Rather it indicates that they aren't interested in competing with IBM mainframes, blade servers or Linux clusters. Since they don't have a credible platform why bother.
There is a strategy here, it is just that none of us can see it clearly yet.
So what's the future of OS X Server? Especially with no true servers to run it on?
Seems to me Apple should update OS X Server to run on ANY server platform.
It will be interesting to see if Apple continues OS X Server development for use with smaller workgroups where a Mac Mini or a Mac Pro might be viable hardware mated with the GUI admin tools. It is hard to see what benefits OS X Server brings to large-scale deployments like data centers where the admins prefer to live in the CLI anyways.
I can't really see Apple licensing the server software to other H/W manufacturers (like HP or Dell) but if Apple wants to make OS X Server available for large-scale deployments I can see them licensing to system integrators like Unisys who would then be responsible for implementation and support. A company like Unisys or IBM is much better structured to provide the support essential for these types of large deployments. I would not be surprised if this is part of the deal struck between Apple and Unisys.
except that mac mini's have no management capabilities and no RAID or hot swappable drives
All of which add to the complexity of the server. Besides you mis the point of a Mini, you hot swap the whole computer. In many cases the swap is many times faster than servicing a 1U server.
with our HP servers i have console access via my iphone. i can even remotely press the power button on a server via my iphone and mobile safari. they don't even need KVM's
Nice security there!
a 1U HP server can take up to 8 hard drives and 192GB of RAM. you'll need a whole rack full of mac minis for that.
Honestly how many web sites need that sort of big iron? Even a Mini can support multiple sites. For that matter a Mini with a little 64GB SSD can support many networking needs.
I'm not here to discount the need for big iron in many cases, rather you seem to discount or completely dismiss the idea that the majority of Internet server needs are rather simple. Currently a dual core Mini can sleep at around 7 watts yet provide fairly robust perfomance when needed to. The fans on most servers take more energy than the Mini.
by the time you pay more money for power, KVM switches, data center space, more racks, you have wasted a lot of money compared to buying Dell/HP/IBM servers
Nope! You want to believe the big iron systems of the past are the solutions for the future, in many cases they are not. The reason being is that big iron systems are not economical it many different senses. Just look at the space demand of a typical 1U server which are generally huge foot print wise. Your arguements about power are just silly.
Like it or not the future of the server business is in very small and low power devices. Expect to see Minis and i86 hardware replaced by ARM based servers in the near future. It is all about being green and competitive. The data ceneter that can move the most data at the lowest cost wins. Space and electrical power these days are bigger issues than raw server power.
Think about servers built around ARM based cards, maybe 3x8 inches that have your network interface, CPU, flash storage and power supply all integrated on that board. Your entire server might hit 5 watts full load. If you don't like ARM look at AMDs new Bobcat based processors. In essence a mini blade server. Mac Minis are currently a approximation of this approach.
and $1500 will buy me a 5 year 24x7 4 hour warranty from HP. with a mac mini i have to take it to an apple store for simple things like replacing hard drives
Anybody with any sense would be replacing those drives themselves. Honestly you can buy a new Mini plus several replacement drives for that sort of money. Further if you are spending that sort of money on service contracts for every server you have you are wasting a lot of cash.
Too many people are acting like the Xserve is unsuccessful because they do not see it in businesses all over the place. The fact is that the Xserve is most popular in education, K12 and higher education for management (open directory) as well as for Xsan/Final Cut Server installations. The Xserve is a very nice 1U machine, and I will honestly be sad to see it go. It is a reliable product that is built very well and did what it needed to do
So?
Seriously so? If the money isn't there how long should Apple continue to support the device?
. Mac OS X server also offers great web server (wiki/blog), podcast producer, and more. You are all discussing this like this is the death of Apple in enterprise, but Apple does not kill a product line without some strategy in place.
Well I will give you this much, this will have no effect at all with respect to Appple in the enterprise. However they killed this product due to the lack of demand.
This is not because they did not sell the product, as all they would have to do is make the server a great tool for iOS management and deployment and they would sell thousands.
not a remote chance of that happening
Apple will not continue to make a product if they do not feel they are innovating in the space and they do not feel it is part of their overall strategy. A perfect example are the RAID storage systems.
Nope both XServe and XRaid where innovative (and expensive). Both died due to lack of sales and a lack of interest from Apple.
Apple used to make APple branded system, but they then stopped and had Promise make all of the systems since they felt other companies could do a better job.
There is a strategy here, it is just that none of us can see it clearly yet.
It is called dumping the line that is not performing.
And I'm going to jam that into the 2U space in my rack how????
Yet another boneheaded move in a string of boneheaded moves by Apple...
If you buy one you can pretty much jam it wherever you like.
So?
Seriously so? If the money isn't there how long should Apple continue to support the device?
....
Nope both XServe and XRaid where innovative (and expensive). Both died due to lack of sales and a lack of interest from Apple.
It is called dumping the line that is not performing.
Too many people think that Apple is run by a "mercurial CEO" who makes one arbitrary decision after another rather than seeing them as the self-disciplined organization they have become over the past decade.
that picture was from 2008
VMWare and hardware in general has grown a lot in the last 3 years. i bet this is all being transitioned to one vmware cluster if it hasn't been done already
VMWare isn't all it is cracked up to be. Besides even if you accept VMWare you can run it on a Mini too. Lets face it some sites are so low in traffic that you could host several on a Mini through virtualization.
The use of Minis isn't an attempt to have one site per unit from what I can see. Rather it seems to be a credible attempt to get around the issues involved in running big iron servers. Google may be able to afford to design optimized hardware but companies like this can't. A Mini though gives them many of the same advantages in that the platform is low energy usage, effectively compact yet surprisingly powerful for its size.
Take a good look at a Mini motherboard. In many ways it is relatively free of extraneous hardware much like a server motherboard.
All of which add to the complexity of the server. Besides you mis the point of a Mini, you hot swap the whole computer. In many cases the swap is many times faster than servicing a 1U server.
Nice security there!
Honestly how many web sites need that sort of big iron? Even a Mini can support multiple sites. For that matter a Mini with a little 64GB SSD can support many networking needs.
I'm not here to discount the need for big iron in many cases, rather you seem to discount or completely dismiss the idea that the majority of Internet server needs are rather simple. Currently a dual core Mini can sleep at around 7 watts yet provide fairly robust perfomance when needed to. The fans on most servers take more energy than the Mini.
Nope! You want to believe the big iron systems of the past are the solutions for the future, in many cases they are not. The reason being is that big iron systems are not economical it many different senses. Just look at the space demand of a typical 1U server which are generally huge foot print wise. Your arguements about power are just silly.
Like it or not the future of the server business is in very small and low power devices. Expect to see Minis and i86 hardware replaced by ARM based servers in the near future. It is all about being green and competitive. The data ceneter that can move the most data at the lowest cost wins. Space and electrical power these days are bigger issues than raw server power.
Think about servers built around ARM based cards, maybe 3x8 inches that have your network interface, CPU, flash storage and power supply all integrated on that board. Your entire server might hit 5 watts full load. If you don't like ARM look at AMDs new Bobcat based processors. In essence a mini blade server. Mac Minis are currently a approximation of this approach.
Anybody with any sense would be replacing those drives themselves. Honestly you can buy a new Mini plus several replacement drives for that sort of money. Further if you are spending that sort of money on service contracts for every server you have you are wasting a lot of cash.
You could add, 192GB of RAM is much better served in a Workstation with 2 or 3 GPGPUs running Engineering, Physics, Biology, Network simulations, etc.
For those of us who have been running Apple's XServe for years, this news, while not entirely unexpected after Apple discontinuing the XRAID in 2008, is pretty damn bad.
So you suspected for 3 years this might happen and now you are angery!
Mac Pros and (lol) Mac Minis are not an option in the data center.
Niether was the XServe for most businesses.
We've been having an emergency meeting about this
You have to be kidding you had an emergency meeting over this. That has to be the biggest joke that I've seen posted in these forums in years. Seriously the discontinued a model and gave you a whole quarters warning, this is not something that requires an emergency meeting. Not unless you make a habit of building moutains out of mole hills.
and we've decided that playing catch up with Apple's lack of stability and Steve Jobs' mood swings all the time (Classic->OSX, PPC->Intel, Software, software, software and now this) is just not worth the cost any more
Are you ignorant or just trying to pull our legs. The Classic to OS/X transition had to happen because everybody was rejecting the old and brittle OS. Like wise with the transition to i86 which had to happen because of the lackluster development in PPC land. As to software what is wrong there, Apple has delivered a state of the art GUI to build any piece of software you could want. Further third party development is going very strong right now.
and we'll be transitioning to Windows in the next two years, on the client side as well.
I need to find out who you are so that I can write the board to detail your rash and unprofessional behavior. Seriously this type of thing has sunk companies. There is no bit of sound reasoning or justification for making a decision like this so fast.
I'm pretty sad about this,
You sad - BS. It appears that you or someone at your place of work was just looking for a reason to get rid of all Apple hardware.
but the plus side is that we won't have to deal with Apple's (lack of) support anymore, which sort of makes it less painful.
OK Be aware that you are making a big mistake. Microsoft is just about the worst possible choice you could make.
You see what is pathetic here is that many people claim that one issue with Apple in the corporate world is that they don't communicate. Yet the minute they try to do the right thing and pre announce a change in the product line up they get crapped on by organizations like yours. You can't have it both ways.
In any event lots of luck with that Microsift based system! The grass isn't any greener on the other side of the fence. Plus we could have a very very long talk about failed Microsoft software initiatives. In the end I suspect you will find out that your very rash behavior will result in far more problems than simply addressing the server issue in a mature fashion.
OK Be aware that you are making a big mistake. Microsoft is just about the worst possible choice you could make.
No. In most cases MS is the BEST choice in the Enterprise.
MS >>>>>>>>>>>>>>>> Apple
Edit:
Job title
Windows Systems Engineer
http://jobs.apple.com/index.ajs?BID=...CurrentPage=11
Yeah! The cowards! How dare they pick and choose what markets they compete in!
I have no problem with Apple choosing what markets to compete in. I do have a problem as a strong supporter of enterprise Apple, being f**ked over with close to no notice. I also am concerned that a company with billions in the bank can't seem to figure out how to address a potentially lucrative market. Apple's 'enterprise push' seems pretty weak without a server product. What they're actually doing is selling gadgets to large enterprise, and semi-servers to SMBs.
. Apple's 'enterprise push' seems pretty weak without a server product. .
There is no real enterprise push for OSX. There is no reason to use macs and osx in the enterprise and it is just a dream of fanboys (in my opinion). In most cases it would be a stupid decision to use macs in enterprise and probably only fanboys would do it. With Apple you have no planning reliability. There are more disadvantages (expensive, bad support, bad backward compatibility, short security support, glossy, ... ). In the enterprise you are working with programs and not with good looking computers. Operating System? Just a means to an end. And in that case Windows e.g. is the better solution: Windows Server with active directory, Exchange, Share Point Server, Office, Windows, ... with Citrix, VMware and so on...
It is all about iOS, when Apple is talking about the enterprise.
The lights-out management, the dual power supplies perhaps? Let me guess, you don't know what lights-out management is, do you?
Oh man. RAID 0 and RAID 1 are for end users, not for data centers. Most data centers run with hybrid RAID 50 or 60 etc. We're not talking about your home office here.
I don't blame Apple for leaving a market they were competing badly in (mostly for lack of trying, IMO)). It would, however, have been nice if Apple had been a little bit more consistent.
You don't know how dumb you make yourself look with comments like that.
Time Machine is an end user/home user backup tool. It's not something that you would use in a data center, where the requirements are at another level entirely.
theolein, I feel your pain. Very few folks here understand the datacenter environment.
Who needs servers when the Cloud thingy is coming.
Truly astounding density. What is that cloud made of, anyway?
Too many people think that Apple is run by a "mercurial CEO" who makes one arbitrary decision after another rather than seeing them as the self-disciplined organization they have become over the past decade.
I just responded to one guy that seems to have this attitude. He claims an emergency meeting about the drpped XServes resulted in a decision to drop Apple entirely both server wise and client workstations. This is complete garbage as you can't make a business decision like this in one day on limited information.
The proble is he goes on to accuse Apple of doing the same thing repeatedly. As an example he gives the Classic to OS/X transition which is absolute garbage. Classic was killing Apple as a computer maker and everyone in the industry should know that. Further Jobs was able to come in, take control of the whole operating systems mess at Apple and quickly get them on track with OS/X. So over the course of the years developers have had only two GUI toolkits to target. Yet this guy wants to switch his company over to MicroSoft software, where I don't even know anymore how many GUI toolkits there are to target.
I'm actually surprised at the indications in this thread of rash and emotional people in business. Really these people should be flipping burgers someplace. Even scarier is that some of these companies might be publically traded. One thing that is expected out of public companies is fisical responsibility and frankly I don't see how anybody can make a responsible decision in less than 24 hours based on this news.
Oh by the way I really don't want XServe to go away but I do expect Apple to make money. If XServe doesn't do that for them then to bad. When it comes right down to it companies will be replacing these boxes with hardware that costs half as much and probably oerforms 2x better. If they are smart they will, but frankly pricing of server hardware from some companies is a little stiff.
Frankly It looks like half the people in this thread feel so damn special that they think they have the right to demand whatever sort of hardware they could want from Apple. Makes you wonder if they are all democrates.
No. In most cases MS is the BEST choice in the Enterprise.
MS >>>>>>>>>>>>>>>> Apple
Edit:
Job title
Windows Systems Engineer
The reasons are clear your whole system can be easily compromized by viruses, bugs and faulty hot patches. That and the fact the MS seems to have lost its rudder with respect to many recent software initiatives, Silverlight being the most recent.
bugs and faulty hot patches.
Sounds like Apple.
Oh I forgot: Apple is not even patching many problems contemporary.
And Security: I don't know if you heard about WSUS and other stuff.