Introduction
Technology has certainly come a long way. On that we can all agree. Many changes in the IT landscape are helpful and involve better performance, flexibility, or new features. However, as most experienced IT folks have learned, not every feature is useful and sometimes the upgrade is a not good idea.
Many concerns around stabilizing or ring fencing old environments are related to security, supportability, or getting better flexibility can be addressed without ditching and starting over. In some cases, even an upgrade is not needed. There is an old saying by Upton Sinclair “It is difficult to get a man to understand something, when his salary depends upon his not understanding it!” Of course your operating system vendor insists that upgrading is 100% needed. Of course your database or application vendor insists that an new version is necessary. As if they'd ever say the opposite!
Top 10 Reasons to Leave It In Place
- You already paid for it. It's been working fine. It's sunk capital still working for you.
- Replacement costs are high $$$. Migration costs are even higher due to consulting fees needed to get the job done. Keeping what you have is free or much less cost.
- Unused features with security problems can be disabled, fenced, or worked around. Upgrades aren't always the best fix.
- Vendors tend to move from buy-and-own onto rent-to-use. Subscription models decrease your revenue by transferring it to the OS, database, or application vendor. In many cases this results in zero benefit to the client or an actual decrease in operational value.
- Older software tends to run considerably faster on newer hardware than modern or current software. Sometimes a hardware upgrade makes more sense than a software upgrade.
- Your operations staff already understands the software they use. Olympic athletes don't change their gear before a competition. Competency boosts efficiency.
- Migrating to new software often requires new hardware and the costs aren't always obvious up-front.
- Upgrades involve new technology that doesn't always increase security, functionality, or efficiency. Consider the heavy use of Java in Oracle's RDBMS product. It's much slower than PL/SQL in Oracle RDBMS and Java with Swing widgets is (much) slower than native widgets. Oracle Corp doesn't care about that, though.
- Upgrades and migrations take time, energy, and money away from (often more productive) other activities in IT. It generates break/fix activity that otherwise wouldn't have been any time at all.
- Vendors escalate costs for older software and OS support. However, OS support can be transferred to PARSEC (saving cash and often increasing quality of service) and the application vendors rarely completely cut off their customers from support altogether due to version issues. For example Oracle still offers sustaining support on their 8.1.7 database released in 2000.