For many value-added resellers (VARs), hardware sales have been in a steady decline. The increasing popularity of server virtualization has compounded the problem.
To increase hardware sales, one technique you might try is pushing Exchange Server 2010. It may not initially sound like an ideal way to boost hardware sales, but Exchange 2010 is a distributed application, and organizations wanting to run Exchange Server 2010 in an optimal manner will have to deploy multiple servers.
I will be the first to admit that smaller customers can get away with single-server deployments and, in many cases, it is even possible to run Exchange 2010 within virtual servers. But this doesn't mean you have to stand back and watch your hardware profits evaporate. Exchange 2010 high availability may be the feature you need to improve your bottom line.
High availability is nothing new for Exchange. It has been possible to cluster Exchange mailbox servers for many years, but Microsoft completely redesigned the way Exchange 2010 high availability works.
In the past, many customers shied away from clustering their mailbox servers because of its complicated nature. More important, Exchange 2007 required VARs to decide up front whether to cluster a customer's mailbox server. If you opted to create a nonclustered mailbox server, you couldn't go back and cluster that server later. Your only option was to create a separate clustered mailbox server and then migrate the customer's data.
Exchange 2010 gives you the option of clustering mailbox servers on the fly, but that alone isn't enough to boost your sagging hardware sales. There are a few architectural changes that can go a long way toward helping you sell server hardware. First, databases are no longer server-level components; they now reside at the organization level. Next, mailbox servers can now be grouped together in database availability groups (DAGs).
How do these changes affect VARs? Exchange Server databases are now server independent, which means they can reside on any server or any combination of servers within a DAG. In other words, it's possible for two servers within a DAG to host the same database. If one server fails, then the other server continues servicing database requests.
Initially, this probably doesn't sound much different than an Exchange 2007 clustered mailbox server. But the kicker is that a single database can reside on up to 16 different servers within a DAG.
As was the case with Exchange 2007 clustered mailbox servers, only one database copy is active at a time. The remaining copies are treated as passive, but you are free to mix and match databases. That way, you don't have all of your active databases residing on one server while other mailbox servers are idle. By placing a mix of active and passive databases on a mailbox server, you will ensure that the workload is distributed in an efficient manner.
Additionally, not every server in a DAG has to host every database. You can have six copies of one database and 16 copies of another. You might have one mailbox server that only hosts two databases, while another hosts 10. The overall distribution is at your discretion.
Using a little imagination, you can put the scalability of DAGs to work in improving your hardware sales. Microsoft markets DAGs as an Exchange 2010 high-availability feature, but by mixing active and passive databases on a server, you can use DAGs as a way of distributing Exchange Server workloads according to each server's available CPU resources. With a little creativity, you can come up with a deployment plan that arranges Exchange 2010 mailbox databases in a way that benefits both you and your customers. You might even be able to talk your customers into purchasing additional server hardware that hosts an offsite copy of their Exchange data.
You may find that some customers like the idea of distributing mailbox databases across multiple servers but that they want to buy low-end server hardware. It is important to remind your customers that the servers within a DAG still need to be treated as clustered mailbox servers. In other words, if a failure occurs, a passive database copy will become the active copy. As such, each mailbox server needs to be equipped with adequate hardware resources.
Typically, this entails equipping servers with high-speed disk arrays, multiple processors and plenty of memory. You need to remember that the more databases hosted on a server, the more hardware resources that server will require.
With this in mind, I will give you a word of caution: Exchange 2010 is the first version of Exchange that can be run either on premise or in the cloud. If you convince your customers to adopt Exchange 2010, there is a chance that they may decide to implement the cloud services model, which translates to very little revenue for you.
About the expert
Brien M. Posey, MCSE, is a five-time recipient of Microsoft's Most Valuable Professional award for his work with Exchange Server, Windows Server, Internet Information Services (IIS), file systems and storage. Posey has served as CIO for a nationwide chain of hospitals and was once responsible for the Department of Information Management at Fort Knox. As a freelance technical writer, he has written for Microsoft, TechTarget, CNET, ZDNet, MSD2D, Relevant Technologies and other technology companies. You can visit Posey's personal website at www.brienposey.com.
This was first published in August 2010