In researching a story for SearchWindowsServer late last week, I came face-to-face with some horrifying numbers. The first is a set of statistics from various sources that indicate more than 10 million servers are still running Windows Server 2003 in production mode in companies and organizations around the world (see, for example, these discussions in Redmond Magazine and in Greg O’Connor’s AppZero blog). The second is the looming date for end of life for that same software on July 15, 2015, where EOL is defined as the “day after Microsoft terminates extended support” for that OS version. You can look this up for yourself at the Microsoft Product Lifecycle Search page, where keying in “Windows Server 2003 R2 Standard x64 Edition” produces the following results (remember: the Extended Support End Date precedes EOL by one day):
I deliberately focused in on the most popular WS03 version to produce a tiny table.
Searching on “Windows Server 2003 R2” instead will cover the whole product family.
Turns out there’s lots of work to do to prepare for a server migration — some of the most important aspects of which I’ll document in my upcoming SearchWindowsServer article for Techtarget (I’ll provide a link to that story right here as soon as it goes live) — so I’m simply stunned to realize that somewhere around 10 million servers in need of migration are still up and running some version of Windows Server 2003 right now.
If this applies to you or your organizations, it’s past time to get going on migration planning. Even with the end-of-year holidays almost upon us, somebody needs to get to work immediately on planning for this effort. The biggest stumbling block is likely to be application compatibility, according to those companies, organizations and enterprises who’ve already been through the exercise. With seven months left to go before the EOL date hits, that doesn’t leave much time to analyze compatibility issues and implement changes, workarounds, or replacements for the applications that so often provide the rationale for using servers in the first place.
If there’s a silver lining to this story, the necessity for change comes with two powerful potential improvements. First, it makes sense to rationalize and consolidate physical Windows Server 2003 server installations in some kind of virtualized form (which means some kind of hypervisor based virtual server environment, or some kind of virtual container for same). Second, it may also make sense to move those virtualized (and consolidated) servers that survive the migration process into the cloud. This will involve considerable work, certain expense, and solving numerous interesting and perhaps even challenging technical problems. But with the end of Windows Server 2003 now clearly in view, hopefully migration will also provide the opportunity to improve and strengthen IT operations along the way.