Xserve - A week of uncertainty

Question

A week since the announcement and I see no clear guidance from Apple. My contemplative confusion is turning into apprehension for the loss of what we had and a sense of uncertainty about the products that my business thrives on supporting. Is Apple not aware of how Xserves are used by their customers? Is Apple not aware of the gaping hole they have created in their product line? Or far worse, is Apple completely abandoning OS X Server and in turn the pro market?
I am not a doom and gloom type of person but the impact that this decision is having on my business and on my customer's perception of Apple could drive anyone to the brink of despair. Let me share with the community some of the conversations that I had this past week.
*Customer 1: North America's largest manufacturer of residential and commercial roofing materials*
_What they have:_ 12 TB Xsan supporting three Final Cut edit bays with a StorNext client in the data center for integrated backup to a CommVault backup system. (I should point out that we also support a graphic design department as well so there are more than 3 Macs in this company.) The Xsan is built on two G5 Xserves and 2 Xserve RAIDs with integration help from internal IT, including the fibre channel department, networking, backup, and others. The solution has enabled the video department to dramatically increase workflow, speed encoding through the use of a qmaster cluster, and improve backup integration with the enterprise CommVault system. The Xsan solution revolutionized the workflow of the department in so many ways that it is difficult to quantify.
_What they need now:_ To expand the volume. They are running over 90% capacity (yes, I know it is dangerous). The Xsan deployment above was done as a proof of concept as it took existing hardware (previously used as departmental file servers) and turned it into a now business critical solution. They have filled the volume in less than 8 months and are now begging for more. The proof of concept resulted in proof that Xsan is a superior product in providing services and features found no where else in the video market. Xsan, with its speed, function, integration with other products, and Spotlight makes the product indispensable. The integration into the corporate Active Directory domain and the ability to collect systems into a render cluster are simple and effective.
+What do we do?+ The existing Xserves are G5s so we are stuck at Leopard and Xsan 2.1.1. Now we have a real tough decision ahead of us. Do we buy Xserves and run with Xsan despite the announcement? And if we do, what is the future of Xsan beyond Jan 31 without a rack mount server. In this customer, rack space is a pricy commodity. Every U counts and I can not consume 12 U for 3 people, regardless of the benefits. Apple has a product, both hardware and software, that satisfies this customers need so well that they are begging for more of it. But can I give them the solution they are begging for? Can I sell them a couple Xserves and run the solution for the near term? But what are my future prospects? What do I do in three years or less from now when they run out of room again? What is my roadmap in the absence of true server hardware? Is there even a replacement for Xsan? Sure MetaSan is looking interesting... but no Spotlight. Sure I can run a StorNext solution... but no Spotlight. We can move to Windows storage, but then I am back on Ethernet file services and no Spotlight instead of a shared file system. If I stand on this hill and say "go with the Xserves as it is the best we have right now" and then there is an announcement from Apple about "one more thing," then my reputation is in question.
*Customer 2: NYC Advertising Agency that is 99.99% Mac based, including server architecture*
_What they have:_ We walked into this customer for our first service call back in February 2004. We found the G4 Xserve we sold them in October 2002 mounted on a wall by another consultant (oh the horror). Within two visits, I had closed all the security holes, implemented OD and single sign on, and started a relationship with the customer that grew from 12 employees, 1 G4 Xserve, and 10 Macs to 120+ employees, 3 Intel Xserves, and 120+ Macs. The Xserves support traditional print design as well as a video, new media, and development departments. Open Directory powers the organization and all Mac clients, over 120 systems, are joined to the domain and enjoy single sign on, centralized management via MCX, and unified access across all services including VPN, groupware (through Kerio), and more.
_What they need:_ This customer is growing. We just upgraded their network backbone to 10GigE, replaced the tape library with an LTO5 40 slot, and added another Active array. They expect 160 users by the end of 1Q 2011. This equals more Macs, more storage requirements and more reliance on OS X Server and the Xserve. In addition, we have had the discussion with them regarding Xsan for the video department. This was a very compelling solution that they were taking a serious look. But now, they are shaken by the announcement.
_What do we do?_ And I ask the community.... How would you respond when this customer asks, "Does this mean we have to switch to Active Directory?" Yes, they really asked me that question. My answer was that the announcement in no way devalues what they have. But I do understand their concern for the future. What do they do in a year? How about 2 years? This is a company running COMPLETELY on Apple hardware and software solutions. Where do we go from here?
*Customer 3: Fortune 500 Retailer of Domestic Merchandise and Home Furnishings*
_What they have:_ In January of 2003, we installed the first Mac in this company. A single G4 tower running Classic applications because the "patient 0" demanded a Mac as part of her accepting employment. Today, this company has over 100 Macintosh systems and three Xserves in three different states. Macs are in the IT department and we went from the platform of distrust to a platform that is a peer of Windows and Unix, including at the server level. The first Xserve was deployed in 2005. The second was installed in 2006. And the final one in 2008. All three are still in active production today and provide critical file services for over 100 Mac users and many Windows users. The servers are all bound to Active Directory, participate in enterprise backup, are monitored by internal IT via SNMP, and provide unparalleled uptime and performance.
_What the need now:_ They are running out of space. No, they ran out of space a long time ago. In October we started the process of proposing upgrades to both the Xserves and the storage to greatly enhance the solution and expand it to cover more departments and likely more Macs. Two of the servers are G5 Xserves so Snow Leopard is out of the question. They were planning on buying two new Xserves and new storage. But as of today, the talks are off. It appears that we are going to lose this business to a Windows storage solution. Farewell Mac servers. Farewell Spotlight searching. Farewell simple management of the Mac systems.
_What do we do?_ What is our argument here? We have none. In light of the announcement, I no longer have a viable solution to offer this customer. The Mac Pro is not a solution. They need more storage and they need a device that fits deployment requirements. The Mini does not do this? How do I get 10+ TB of storage connected and how do I support 100+ concurrent users?
*Customer 4: US Branch of Major Magazine Publisher that is 99.99% Mac, including server architecture*
_What they have:_ This customer is the poster child for Xserves. They have 11 Xserves, including a G4 still in active use. At their peak, we were running 16 Apple Xserve RAIDs (these have mostly been replaced by Actives). These Xserves run file services, FTP, Mail (Kerio), automation, Open Directory, offsite replication, and much much more. And here is the kicker! We deployed them in May 2006 running 10.4.6! The last time the primary cluster was touched was January 2009 when we upgraded to 10.4.11. Yes! There are 7 of the 11 servers still running 10.4.11! And do you know how often we service that cluster? 4 times a year. That is it. There is no downtime. There are no problems. This is an environment of about 450 users (all Mac) working directly from servers on multiple shifts. All machines are bound to OD and we are using MCX, Kerio Mail Server, UTS, NetInstall, and much more in this environment. And the whole thing is managed by a staff of 2 technicians. Oh, and in 2007 we built a dedicated datacenter for the Xserves, including proper power, cooling, and fire prevention. This set the foundation for a 10 GigE backbone and all servers operating with aggregated links to stacked edges switches with EAPS ring failover. This environment is a model of enterprise deployment of Apple hardware rivaling any Windows deployment out there.
_What they need:_ The customer is currently rolling out workstation replacements company wide. These units are 10.6.x. The 10.4.11 server are doing fine but it would be nice to unify and get all the MCX stuff working without the hacks. Purchasing new Xserves was a planned project for the beginning of next year.
_What do we do?_ This is the one customer in which we will likely still push the Xserve before it is no longer available. The nature of the business and the cyclical nature of upgrade cycles can likely allow us to operate on 10.6.x at the client and server for at least 4 years. So despite what Apple is doing, we may be able to keep the dream alive in this one customer. But in February will we be standing with egg on our face for a foolish recommendation?
*Customer 5: Fortune 100 Insurance Company*
_What they have:_ We support Mac departments in 4 states. Two of the departments are running Xsans. These are obviously driven by Xserves that have been embraced by corporate IT. In corporate America, projects take time to plan and approve. The Xsan expansion project has been in the works most of the year. It has been approved is now ready to start placing orders. But we are clearly caught in a tough position.
_What they need:_ More storage! And InnerPool from Active. InnerPool will go down in history as the greatest product that never will be. The success of Xsan in these two departments almost can not be measured. The undeniable reality is that the solutions is a wild success and the departments want more.
_What do we do?_ The word from Active is that InnerPool is dead and will never reach a customer. The Xserve is no longer. Yes we can buy it and yes it is approved. But I have the same uncertainty as listed above. Great, I deploy Xserves and Xsan today. But what is its future? And without InnerPool I has inefficient LUN carving.
Conclusions:
I am sure that my scenarios are shared by many in the consulting and IT fields. Over this past week I've struggled to understand Apple's decision to discontinue the Xserve. I've struggled to ease the concerns of my customers. I've struggled to find the correct message to tell customers in this time of uncertainty. I've struggled to understand what my company will do to fill the void left by the absence of the Xserve. And, most troubling, I've struggled to understand what the loss of the Xserve will do to Apple's enterprise solutions including the future of OS X Server and the traditional markets Apple supports.
I understand that on the surface this is the retirement of a hardware product. I understand the basic argument that the product was not a powerful seller. But being an Apple Reseller and ACN, I also know that no one is pushing Xserves other than us. The Apple retail store is incapable of pushing the Xserve as a product. And please, real businesses are not going to Apple retail. Does anyone walk up to a person selling iPods and have a discussion about real business needs? And why was I never able to find the Xserve on store.apple.com's home page? If you want to sell a product, you really need to let your customers know it exists.
I understand that Apple is all about the one thing. They always have been. When the iMac saved Apple it was the hot item. Then the iPod replaced it. Then the phone. Now the iPad. Apple focuses like a laser on the one thing that is big. But what they are missing is that there is a loyal group of customers in content creation that are only able to do their work because of the power, ease of use, and products that make up the "pro/enterprise" segment of Apple's business. This includes the Mac Pro, the Xserve, OS X Server, Xsan, Final Cut, and more. What is the future of all these products? The Mac Mini Server is surely not the future.
Again, this is speculation driven by lack of guidance and leadership. But if my customer reaction is an indication, the inability to deploy a server class device means the inability to run OS X Server. This will mean that all of the duties found in OS X Server will be replaced by other systems. This means no OD, no AFP, no NetBoot/NetInstall, no iCal, no Xsan, no qMaster, no MCX, no synchronized home folders. The list goes on and on. And by translation, this likely means the decline of the viability of Final Cut deployments and even content creation in general. If we are going backward to isolated workstations and the inability to deploy space efficient direct to broadcast solutions, the business will be going elsewhere. In the space of a single Mac Pro, I could deploy and entire 50 TB Xsan with 1U left over (2 Xserves, 2 Active arrays, 2 FC switches, 1 MD switch). Apple is not offering progress and this is not a better solution.
When Apple retired the Xserve RAID, Apple provide the community with a better solution. The Xserve RAID was a product of its time and time marched on. Apple was not interested in maintaining a peripheral product and rightly left the business to organizations specializing in mass storage. By bowing out of the segment, the community got faster devices with more redundancy. The difference then was that Apple said, "hey, we are not going to make the Xserve RAID anymore but here is something better." Today, they are saying, "hey, we are not making the Xserve any more and we have nothing better to tell you."
Yet, the Xserve RAID was a peripheral. It was not a host system running Apple's OS. For the company that is known for making the whole widget, how does this announcement sit? We are now missing the hardware component of the whole widget and the options, virtualization or licensing to third party hardware vender, goes in direct grain to everything Apple has done.
I know others in the community have covered this. The hardware Apple is offering is not a replacement. Honestly... A Mac Pro that has a hardware problem is a minimum of 2 days of lost operations for a company. Take the power supply. On an Xserve, you have a second one and if you were smart enough to get the spares kit, you can replace it in less than 1 minute with one hand tied behind your back. No downtime. No service disruption. Now, the Mac Pro... There is no spares kit. That means if the unit goes down you call Apple to diagnose the problem (or good golly you bring it to the Apple Store?), they send the part to your closest Apple Authorized Repair Center, and then they arrange for an onsite repair. This process will put you out of business for 2 days minimum. Unacceptable.
I am not sure who Apple thinks their server customers are. But the customers I support will never be satisfied with the Mini Server. Apple targets the Mini Server to SMB. Well, let's assume that the SMB has 10 iMacs. The iMacs have a collective storage capacity of 10 TB. The Mini has 500 GB. How does that help the environment. If we look at Apple's traditional customers running content creation, a single user could fill the server with 100 photoshop files or a few video files. And don't get me started on the expansion issues with the Mini.
Apple, regardless if the end result is licensing server to another vender for hardware development or allowing virtualization, please open up the communication channel and reassure a shaken industry on what you see as the realistic future of your server products. The guidance of your transition guide is not guidance at all. Please come down from the iPod cloud and understand that you have a lot of customers quavering in uncertainty.
And finally, if Apple is using the retail stores to provide guidance to the types of businesses that they sell their product to, then they should call us. We have been integrating Macs in enterprise for over a decade. We have Macs in every pharma in the state, many in regulatory roles meeting strict government requirements. We have Macs in insurance, research and development, finance, law, and more. These system meet or exceed HIPPA, SOX, FDA, FTC and other government regulatory requirements. Apple, you are already in the enterprise. Please don't give up on what has taken years to achieve.
Thank you for reading. We are passionate about the services we offer and strive to give our customers the best possible solutions. Apple has enabled us to do that for a number of years. Now I am concerned. As an ACN employing numerous ACSAs we are concerned with the writing on the wall. Apple, please provide more communication on your intentions.

Answer 1

Funny you should ask... At about the same time that the Xserve was discontinued, we also lost our channel rep of nearly 14 years. Right now we are not sure who our rep is. I have however, sent this to as many people in Apple that I could think of. And I plan on continuing to reach out to as many as possible.
The sad reality is that this decision is impacting customer deployments and the future of customer installations. Apple strives for market share and loves to tout those numbers. But the numbers we fought for over these last eight years are going to disappear.
As much as this impacts our business, we are making up for it in enterprise deployment of systems. We will survive and lead in our geographic area. I do not fear that. But OS X Server is a UNIX server and with it has come all of the benefits of a UNIX system, including legitimacy in a data center. The Intel Xserve cemented this reality because of its hardware configuration and form factor. Yes, we have G4 and G5s still in service. But those are mostly in all Mac shops that understand the benefits and can work around the buzz word limitations. But fortune 100/500 IT departments did not really get on board until the release of the Intel server.
Still beating the drum to raise awareness. Spread the word and talk to your Apple reps.
Also, ask about the direction that Apple's Enterprise team is going. If the news I hear is correct, then our fears about all things pro and enterprise going away are likely true.