There is no turning back the clock-─April 8 soon will be here and, with it, the discontinuation of Microsoft’s full support for Windows XP. While many organizations have completed the inevitable migration to Windows 7 or 8, many more have not. Time is growing short for those organizations, but skimping on planning and preparation for the migration will only bring more headaches and more costs downstream. Each phase of the planning process—application discovery, rationalization, assessment, remediation and User Acceptance Testing (UAT)—is crucial to reducing cost and corporate risk, but we have found most organizations that experience a migration failure have failed to adequately complete the first step: application discovery.
Companies often underestimate the major obstacles caused by application compatibility issues during and after a Windows migration. Application compatibility is the biggest obstacle for enterprise Windows 7 and 8 migrations, but it can be overcome with automated compatibility testing to determine application readiness and improve not only the scoping and planning of the effort right from the start, but, critically, the quality and speed of the implementation. Running an automated assessment against real application data gives incredible insight into the potential starting point—without this, there can’t really be any confidence in what the expected end will be. Many have found that it isn’t as bad as they think. It’s all about eliminating the “known unknowns.”
Organizations that overlook this important step, or don’t commit enough resources to it, likely will find themselves knee deep in a migration project that is not only running overtime, but brings with it costly post-migration failures such as user downtime, missed business deadlines, loss of competitive edge, lower staff morale, and decreased stakeholder confidence.
Organizations have told us that some of the post-migration problems they’ve experienced from neglecting to plan and test properly through a migration project include business-critical applications that don’t run correctly, application crashes, the need for emergency fixes, last-minute additions to the migration list, user downtime, and a myriad of others. All of these issues stem from diving into a migration project without a thorough review of the existing users, groups and applications used, as well as how critical those apps are to running the business, and without performing an in-depth application compatibility assessment and remediation to avoid known issues.
Whether your migration is still in the planning stage, or even if you haven’t yet begun, don’t panic. Your migration ultimately will be the most successful with planning and preparation that includes effective application discovery and thorough application assessment. Not only will your application portfolio be simpler to manage after the migration, but application readiness will lead to a higher rate of rationalization and resulting cost savings. Ultimately, that means a better user experience and a much faster migration or adoption of technology change.
The key to cracking the compatibility challenge and facilitating better application management is automation. It’s very difficult to manually test your applications to make sure they will install, run, update and uninstall correctly without breaking your operating system or other applications. It’s also hard to manually test application dependencies on middleware and the OS version, and to confirm that the apps will not conflict with each other. Implementing a knowledge-based system that automates the testing of hundreds—or even thousands—of applications at a time against a set of comprehensive rules to determine whether they will work on a new target operating system or technology will capture issues up front. As a result, you will save the time, expense and complexity of reversing application compatibility failures later, and, most importantly, you will avoid user disruption both during and after the migration.
Once a new compatibility issue is discovered, adding a new “rule” that searches for that pattern is essential to keeping your system up to date. Finding a knowledge management solution that not only provides regular rule updates to keep up with current platform checks, but also allows you to create your own rule checks to capture issues that may be unique to a particular environment is crucial. In other words, you should be able to run queries against all the metadata collected from your application estate to search for potential issues. A good knowledge management system will ensure that you can benefit from a continuous feedback loop.
Don’t even think about putting off starting your Windows XP migration project after April 8. If you do, security will be your biggest worry. Even though Microsoft will continue to provide updated anti-malware/virus security signatures, there will be no fix provided if there is a breach unless you “invest” in costly extended XP support. If you have a lot of critical applications but don’t have access to patches, your network has taken a leap in vulnerability.
Application readiness should be viewed as a true investment in ongoing management, rather than a one-off problem for a particular migration project. Migrations to new Windows operating systems happen more and more frequently, and Microsoft already has started talking about Windows 9, which is rumored to be available in another year. The whole concept of migrating or transforming your environment isn’t going to go away. New applications and new versions of apps need to be introduced into the environment on an ongoing basis, so application readiness and agility will put you in a much better place when it’s time to migrate to Windows 9, and whatever comes after that. Migrate successfully now, and you can future-proof for the next one.
Ann Maya is the Senior Product Manager for Dell ChangeBASE. She has held various software development, UNIX data, application management and product management roles for companies including Hewlett Packard, Compaq, Morgan Stanley, App-DNA and Quest Software.