[hmurchison2001]

I personally have never been a fan of the theory that we need Mac clones. I think it’s a faulty assumption that the roadmap Microsoft took to success could somehow be replicated by Apple via cloning the Mac OS.

However I don’t believe this ideology extends to the SMB and Enterprise markets. I believe there “is” money to be made in supporting non-Apple hardware in this environment. The chief factor in driving this change is hypervisor virtualization. I routinely attend conference calls with our sales reps and their clients and I cannot remember the last time I heard and IT guy say his company was not interested in Virtualization at any level.

My belief is that Apple handles their desktop/notebook/workstation lineup very well with frequent refreshes (for the most part) and excellent design. Apple attempted to bring this same panache to the business sector and I believe they were successful. The Xserve RAID and Xserve were/are nice pieces of technology and look gorgeous. But as we’ve seen recently Apple’s interest in refreshing these product lines is more tepid than their consumer offerings. The Xserve RAID is dead and a RAID box from Promise has taken its place. The Xserve has evolved but it isn’t the server for everyone IMO (dearth of drive bays and high entry level cost).

My thoughts are Apple should just jettison the Xserve and focus their efforts on partnering up with a few high volume vendors and beefing up their support. For instance if Apple partnered with VMware for ESX support for a hypervisor version of OS X Server and a high volume vendor like Hewlett-Packard they’d have access to amazing technology from both companies at a fraction of the cost of engineering their own hardware.

Suddenly OS X Server is sitting right along Red Hat, Windows Server and more in a Rack n Stack or Bladed virtualized environment.
The benefits are clear:

Pros- no need to engineer hardware. Access to a huge lineup of hardware products. Halo effect as more OS X servers means more potential iPhone/desktop/notebook/workstation sales. High Margin service and support sales.

Cons- driver development and application support. Apple would need to work diligently with VMware and HP to make sure they support a good amount of hardware and software tools.
Today even at the server level hardware is a commodity and it’s the software applications and management tools that make the difference.

Apple is a software company yet they “need” the hardware profits to fuel R&D at the consumer level. They aren’t making enough money with the Xserve to really care about “owning the whole widget”. Divest the burden Apple and partner with those already entrenched. You win either way

[besson3c]

I agree that Apple's server offerings are lacking, and that the wisdom that holds true with their consumer products does not also true with their server products. I agree that licensing OS X Server and getting OS X to work in a virtual machine is in Apple's best interests.

However, I also think that OS X Server itself needs a lot of attention if it is going to be of value in enterprise environments. As of right now, my thinking is that Apple is trying to position OS X Server as a small business offering/just-add-water sort of server that would compete against a small business looking at a Dell/Windows combo. However, OS X Server completely falls apart when you put it against Red Hat or Windows Server in more enterprise level environments. It needs a lot of attention...

I could list the areas in which I think it needs the most attention, but I don't want to derail this interesting thread!

Edit: this inspired my own thread: http://forums.macnn.com/90/mac-os-x/...s/#post3661154

[Randman]

Apple ain't gonna license anything like that for a looooooong time.

[mduell]

Is there really a crowd out there that wants to run OS X servers on lower-end hardware? A crowd where the Mac mini isn't enough, the iMac isn't right, and the Xserve is overkill (or not enough)?

I work with a bunch of huge Mac fans and we use Leopard Server on an Intel Xserve at work for some things (Jabber, Wikis, some filesharing, and Time Machine), but even they recognize OS X Server has some severe limitations and we end up using FreeBSD and Linux servers for most network services.

[hmurchison2001]

Besson3c

I know. Leopard Server was a good step in the right direction but the only way for OS X Server to be more than just the "add on" server for Windows/Linux environments is for Apple to invest some resources into it. I don't believe they can invest enough resources into their Biz strategy to close any gap with the HP/IBM/Dell of the world just as those companies will likely never had a good iPhone competitor.

Imagine the breath of fresh air that Apple would smell by attacking the biz sector from this angle. They wouldn't have to worry about giving roadmaps for hardware and maintaining lifecycle charts because like Linux ..they don't design hardware (anymore) thus they just keep up with supporting the must have hardware and deliver outstanding support.

I can tell you right now the the potential for media applications would be fantastic.
Apple has XSAN for SAN Filesystem support and XGRID could actually be useful for servers
located on a fast Blade system backbone. Apple has Final Cut Server now which at the client level is
cross platform.

It just makes sense to divest themselves from an financially underperforming hardware platform (Xserve) and
focus on the software instead which is Apple's forte.

[hmurchison2001]

Originally Posted by mduell 

Is there really a crowd out there that wants to run OS X servers on lower-end hardware? A crowd where the Mac mini isn't enough, the iMac isn't right, and the Xserve is overkill (or not enough)?

I work with a bunch of huge Mac fans and we use Leopard Server on an Intel Xserve at work for some things (Jabber, Wikis, some filesharing, and Time Machine), but even they recognize OS X Server has some severe limitations and we end up using FreeBSD and Linux servers for most network services.

I don't even think it's really about low end hardware. I see so many companies that just want to reduce the costs of managing their server farms. They want to consolidate. They want to reduce the amount of switch ports they manage (especially if they are using fibre channel)

No lets say the graphics or marketing dept are all on Macs. Now does the IT dept in a company looking to consolidate reeeeally want to have to manage that lone Xserve sitting their supporting the Graphics/Marketing dept or are they thinking

"GD Apple making me keep this box here for OS X Server when it's the same damn guts"

It sure would be nice to be able to toss OS X Server right on top of ESX and manage it with the current set of tools.

It's hard to see the downside.

[besson3c]

Originally Posted by mduell 

Is there really a crowd out there that wants to run OS X servers on lower-end hardware? A crowd where the Mac mini isn't enough, the iMac isn't right, and the Xserve is overkill (or not enough)?

I work with a bunch of huge Mac fans and we use Leopard Server on an Intel Xserve at work for some things (Jabber, Wikis, some filesharing, and Time Machine), but even they recognize OS X Server has some severe limitations and we end up using FreeBSD and Linux servers for most network services.

Just what is the OS X Server crowd then? Is it just people that need to operate computer labs of OS X workstations? Small businesses looking for a Mac-based alternative? Creative shops that need a server and don't have a sys admin?

This seems like such a tiny small market I wonder why Apple even bothers?

[besson3c]

Originally Posted by hmurchison2001 

Besson3c

I know. Leopard Server was a good step in the right direction but the only way for OS X Server to be more than just the "add on" server for Windows/Linux environments is for Apple to invest some resources into it. I don't believe they can invest enough resources into their Biz strategy to close any gap with the HP/IBM/Dell of the world just as those companies will likely never had a good iPhone competitor.

Imagine the breath of fresh air that Apple would smell by attacking the biz sector from this angle. They wouldn't have to worry about giving roadmaps for hardware and maintaining lifecycle charts because like Linux ..they don't design hardware (anymore) thus they just keep up with supporting the must have hardware and deliver outstanding support.

I can tell you right now the the potential for media applications would be fantastic.
Apple has XSAN for SAN Filesystem support and XGRID could actually be useful for servers
located on a fast Blade system backbone. Apple has Final Cut Server now which at the client level is
cross platform.

It just makes sense to divest themselves from an financially underperforming hardware platform (Xserve) and
focus on the software instead which is Apple's forte.

I don't know... I think that ZFS is going to blow Apple's RAID solutions out of the water, and I don't see a compelling reason to use Apple's grid computing solutions in a high performance computing environment that is not based around Macs... I'm thinking that it was designed for rendering farms?

It looks like Apple has done a really good job positioning the iPhone for enterprise, so perhaps this is a signal that they will start taking this market seriously?

[besson3c]

Originally Posted by hmurchison2001 

It sure would be nice to be able to toss OS X Server right on top of ESX and manage it with the current set of tools.

Ditto for VMWare Server, Xen, KVM, etc. if you are running a Linux/Unix based shop...

[hmurchison2001]

Originally Posted by besson3c 

Just what is the OS X Server crowd then? Is it just people that need to operate computer labs of OS X workstations? Small businesses looking for a Mac-based alternative? Creative shops that need a server and don't have a sys admin?

This seems like such a tiny small market I wonder why Apple even bothers?

That's what I'm scratching my head about. I'm mean it's create to have an unlimited client OS X server with the $3k Xserve but looking at Leopard server ...there's barely enough tools and middleware apps to support a few hundred people for general computing needs IMO

Originally Posted by besson3c 

I don't know... I think that ZFS is going to blow Apple's RAID solutions out of the water, and I don't see a compelling reason to use Apple's grid computing solutions in a high performance computing environment that is not based around Macs... I'm thinking that it was designed for rendering farms?

It looks like Apple has done a really good job positioning the iPhone for enterprise, so perhaps this is a signal that they will start taking this market seriously?

The iPhone is well positioned. Apple would do well to offers something close to RIM's Blackberry server in the future so that they aren't just depending on Microsoft's Active Sync for push.

Originally Posted by besson3c 

Ditto for VMWare Server, Xen, KVM, etc. if you are running a Linux/Unix based shop...

Indeed. They could start off with VMware or Xen and expand if they find success.

I think the frustrating thing is seeing Apple, a company known for attention to detail, sleepwalk with certain products. Apple's good but they're not THAT good. Either put some effort into OS X Server or vacate the platform and focus on making OS X desktop and mobile platforms the ones to envy...even more.

[besson3c]

Apple seems well positioned to make contributions to enterprise messaging products. Apple is one of few companies that can compete with MS Exchange, and it looks like Zimbra's future is being threatened.

It's not only an Exchange replacement where they could contribute though, but in a push IMAP server or plug for something like Cyrus IMAP, in presence management, in improving and contributing to IMAP and producing rich IMAP clients (they can do a LOT more with OS X Mail), developing groupware standards (iCal Server is a good step in this direction, but it isn't being offered enough resources), coming up with a better way to transfer and store large files (IMAP is not a file transport mechanism, nor was it designed to be), opening up their nice AJAXey web-based client they offer to .Mac subscribers... There is a LOT that can be done here, and all of this could be a great catalyst for selling more iPhones.

[mduell]

Originally Posted by hmurchison2001 

No lets say the graphics or marketing dept are all on Macs. Now does the IT dept in a company looking to consolidate reeeeally want to have to manage that lone Xserve sitting their supporting the Graphics/Marketing dept or are they thinking

"GD Apple making me keep this box here for OS X Server when it's the same damn guts"

It sure would be nice to be able to toss OS X Server right on top of ESX and manage it with the current set of tools.

It's hard to see the downside.

Ohhhh, you're one of those 'virtualization-solves-all-problems' guys. Sorry, I missed that at the start of the thread.
Why did the graphics or marketing department need an Xserve in the first place? What service couldn't be run from existing IT infrastructure?

Originally Posted by besson3c 

Just what is the OS X Server crowd then? Is it just people that need to operate computer labs of OS X workstations? Small businesses looking for a Mac-based alternative? Creative shops that need a server and don't have a sys admin?

This seems like such a tiny small market I wonder why Apple even bothers?

If nothing else, it's a pretty low cost/effort derivative (OS X client + a few pretty GUIs for standard FOSS software) that helps with a sales in a few markets (education?). Some of the limitations (like no periods in email addresses) make it pretty obvious they haven't put much effort into making it attractive for businesses (no firstname.lastname@company emails).

[besson3c]

Originally Posted by mduell 

Ohhhh, you're one of those 'virtualization-solves-all-problems' guys. Sorry, I missed that at the start of the thread.
Why did the graphics or marketing department need an Xserve in the first place? What service couldn't be run from existing IT infrastructure?

I can't think of any service, but of course there is some demand in virtualizing OS X for development/testing/support.

If nothing else, it's a pretty low cost/effort derivative (OS X client + a few pretty GUIs for standard FOSS software) that helps with a sales in a few markets (education?). Some of the limitations (like no periods in email addresses) make it pretty obvious they haven't put much effort into making it attractive for businesses (no firstname.lastname@company emails).

Last I checked OS X Server uses Cyrus IMAP for their IMAP server. In Cyrus, the period is used as a namespace delimiter, e.g.:

user.me
user.me.myfolder
user.me.Junk

You can change this by modifying Cyrus' imapd.conf, but I have no idea whether Apple's GUI would clobber this manual change. That is always the question of the day.

Perhaps OS X Server is low-dangling fruit for Apple, but it also makes them sort of the laughing stock in the world of servers. Apple's marketing department generated hype is legendary, but they have a history of under-delivering in this area (e.g. they used to call their old XServes "enterprise class", the whole "enterprise grade" Time Capsule drive thing, etc.). Apple probably should be more upfront and go after providing real solutions that IT people can confidently recommend, rather than just going after the businessey type folks that are attracted to shiny GUIs, pretty websites, and marketing fluff.