The biggest problem with a wholesale migration off of Windows Server 2003 is that it can turn into a "bet -- and...
lose -- the farm" activity if it is not properly researched, planned and managed. You'll have to solve any application issues, which means a complete inventory, compatibility assessment and tough decisions about how to deal with elements that won't or can't readily or easily make the move. But there are some important elements to factor into your risk management regime.
First, there are early-stage issues with a Windows Server 2003 upgrade related to what could be called consistency and correction. Consistency means establishing or imposing order on the environment that will operate post-migration. It also inevitably means cleaning up and rationalizing applications, data and services amidst the pre-migration status quo. These issues must be fully addressed and resolved before the migration can get underway. Correction refers to rationalizing applications and architectures so that you don't simply move whatever you have, but rather, take the opportunity to rethink and to re-architect the resulting target environment post-migration where necessary. This must be complete before migration can begin and often involves what Texans like to call lots of "cussin' and discussin'."
Next, compatibility issues in a Windows Server 2003 upgrade must be addressed. This is particularly important for applications, data and services that may be used pre-migration, but will have to go through changes or replacements to keep running properly post-migration. Continuity of service is an issue that covers the whole spectrum of activities, starting with migration planning, piloting and testing, through the staging and deployment of the actual full-scale migration and well into the post-migration phase. The key concept here is that data, applications and services must remain available through the process to keep the migration from impacting productivity or profitability before, during and after its occurrence.
Finally, contingency refers to the expectation that problems will present themselves, systems and applications will misbehave. Contingency planning addresses continuity in the face of such problems and may depend on shadowing activity and applications on old systems and services as new ones are brought up and a cutover from old to new takes place. If the shadow remains available, even when the new world experiences a glitch, business can go on and keep going while troubleshooting, repair and restoration take place behind the scenes.
Secure the Windows Server 2003 upgrade
Early on, sandboxing in the lab to analyze and try out migration approaches and tools is essential because sandboxing provides early warning for potential "gotchas." This gives IT the opportunity to find, fix or work around problems in a lower-pressure, non-production setting in the lab before they get serious. For larger enterprises, this approach also lets IT find and practice with tools to perform server virtualization, image capture for backup and VM restart in case of problems with migrated systems and services, deployment and integrity or go/no-go checks well before dealing with production systems and services.
Among other things, sandboxing enables the IT deployment team to develop and document processes for the migration as well as observe and establish go/no-go strategies during planned deployment or update windows. IT must make sure it understands what needs to be done, how long it will take and what kinds of experts should be on tap for troubleshooting.
Of all the elements that can make or break the migration, it's crucial to create a library of canonical images of systems prior to migration that are ready to be fired up as VMs; you must also find a way to gain access to those images for backup and file recovery maneuvers. Many vendors, such as Acronis and Unitrends, offer backup and VM management tools that can capture server images before the fact and later restore those images to bare-metal machines where physical servers make sense or VMs where virtual servers are needed. These libraries can also help establish and maintain an important safety net that a well-orchestrated contingency process will provide during and after the Windows Server 2003 upgrade.
About the author:
Ed Tittel is a 30-plus year IT veteran who's worked as a developer, networking consultant, technical trainer, writer and expert witness. Perhaps best known for creating the Exam Cram series, Ed has contributed to more than 100 books on many computing topics, including titles on information security, Windows OSes and HTML. Ed also blogs regularly for TechTarget (Windows Enterprise Desktop), Tom's IT Pro, UpperTraining.com and PearsonITCertification.com.
Three questions to ask before you move from Windows Server 2003