Apple killed Xserve and, worse, seems to have no plan.
Apple’s killed Xserve. They’ve pledged support for the service life of the machines currently in the field or bought before January 31st 2011. They’ve pledged OSX Sever support on Mac mini and Mac Pro. They only mention Snow Leopard Server no mention of Lion Server which is, in silence, damning.
http://images.apple.com/xserve/pdf/L422277A_Xserve_Guide.pdf
They’ve made a very, very bad move.
It’s pretty obvious from this move that Xserve was not a high volume or ‘successful’ product unto itself. That success, measured only in revenue and margin of the XServe product line was likely pretty bleak.. The problem is, sales, profits on XServe as a product is not how Apple needs to measure XServe’s value to the company.
XServe was, in part, a means of justifying a place for Macs in corporate I.T. That the product existed said something corporate I.T. needed to hear from Apple in order for the Mac to be allowed at work. I do mean allowed. If you have never worked in a corporate environment, dealt with corporate I.T. thinking, never been the ‘guy who supports those weirdo designer and video people’, you have no earthly notion of the proactive steps certain sorts of people will take to get Apple products off their campus by any means necessary. And yes, even now, maybe more now, with the consumer-focus of Apple’s balance sheet.
The current model XServe is actually, in small part, the result of a ‘listening tour’ Apple did with some corporate clients including a former employer of mine, One of the features loudly advocated for by a colleague was Lights Out Management (http://support.apple.com/kb/HT2773) and Apple responded. Surely not just to his feedback but to what was likely feedback they got from many many other similarly able sys admins like my then colleague. The issue wasn’t so much how many potential clients would actually use lights out management but how many of those with purchase approval power would perceive that, by including LOM, Apple was serious about MacOS in the Enterprise.
More to the point than that was Apple’s investing the resources, the staff, in making the trip to what was a major customer and making them feel heard. That my colleague saw his suggestions show up in a new product helped Apple begin to regain their foothold there. Apple sold a lot of iMacs and Mac Pros to that company since then. By a lot, I mean significant hundreds and this was hardly a large company or even a shadow of a major higher education site.
In canceling Xserve, and presumably and more importantly scaling back the staff involved dedicated to supporting and evangelizing OS X as a server platform and full peer in business. Apple isn’t just shedding an unprofitable and tangential product, they are jeopardizing sales to academia, research institutions and medium sized business. In doing that, they jeopardize sales of all Macs and from there, iOS devices across all markets.
Killing XServe without a forward looking indication of Apple’s commitment to I.T. in business is a capital B bad thing.
I.T. people, if they’re actually competent I.T. people, are a paranoid bunch. We like contingency plans. We like backup. We like automation. We like fault tolerance and high availability. 5 Mac minis without an Apple supported clustering and HA solution is not a viable solution to an XServe with duel power supplies and fault tolerance on disk. A Mac Pro consumes absurd volumes of space in some of the most expensive real estate in the world. A fire-suppressed, raised floor, redundantly cooled, UPS and generator powered machine room could be the most expensive, by cubic inch, space in the world.
By killing XServe, Apple is saying they don’t understand and are not interested in I.T. people. The problem with that? If Apple’s not interested in I.T. people they won’t be interested in Apple and all the gains Apple has made as a respected peer platform in mixed OS environments are in jeopardy.
If a company’s I.T. department won’t support a Mac on their LAN, they won’t support a Mac used to connect via VPN from home. Their users will be right back to “I can’t buy a Mac because I can’t use one at work and I can’t even get permission to use mine at home to do my work from home. I have to buy a PC.”
As much as Apple may want iPad and iPhone to be their primary high volume bread winners, they can’t afford to kill the Mac. They need the Mac to be the development platform for iOS. They need the Mac to be the “one sure bet your iOS products will work smoothly with” for customers. The potential community for developers can’t be left to hope iOS dev tools will install on Windows. Apple needs the Mac and the Mac still needs to be a viable computing platform, not an appliance. That means a developer needs to want to use a Mac. The Mac must stay a viable platform. The only way that happens is if Apple invests just enough to insure the Mac is a full peer platform for business use.
Apple can mitigate some of the costs that prevented XServe hardware from being a viable product by expanding the market for the hardware and throwing away the cost of an additional logic board design. They can reach a market under-served by the incredibly deep and somewhat physically delicate XServe 1U design by rethinking the Mac Pro.
A new ‘Mac Pro-Rack’ and commitment to the OS X Sever software platform.
Mac Pro-Rack is a variation on the current damned near perfect enclosure design of the Mac Pro. It is, now, no taller than it it could be wide to slide into a rack (with the handles removed). There appears to be very little about the internals that demand vertical orientation for proper cooling. The optical drives could be rotated 90 degrees at the, worst case, expense of access to one slot PCI slot. Dual Power supplies, there’s room for or make an external second supply with DC inputs an option. Hell, make Mac Pro-Rack a bolt on kit for a redesigned horizontal or vertical Mac Pro. When used as a workstation? Smooth top and bottom, front and rear facing (rather than top and bottom) handles. When used in a rack? Bolt on optional ears and rails after removing a piece of sexy Apple-designed trim.
No, clearly it’s not literally this easy. ruggedizing for use in a mobile rack, thermal management and rack hardware of course demand engineering resources but the point is, it’s possible for Apple to have a shared logic board design for their pro workstations and a rack mountable product and mitigate the development  costs of ‘specific for use as a server’ hardware. A good 4U rack mountable Mac Pro is of value to use in an I.T. server rack and in media production markets. (and military, security and field research applications as well).
Build this dream hardware or not, Apple can not afford to simply dump their visible interest in supporting Apple in enterprise environments. It may be a cost bit but, without visible effort on Apple’s part? Again, Microsoft, Oracle or others will be able to edge the Mac out of a place on the corporate approved purchase list. Apple needs a continuing story about the suitability of the Mac as an all purpose platform.
Apple won’t make money selling Xserves. They will make less money selling Macs without a product in XServe’s niche. There needs to be a visible plan and commitment. The will make less money selling iOS without the Mac and, long term, without the Mac, iOS is doomed. iOS is not a ‘making things’ platform.
They should have bought Sun.