Upgrades occur for a number of reasons. Hardware or middleware software goes end of life. Going to a new operating system is another common upgrade. Applications need upgrading to provide new functionality. One aspect of upgrading is trying to understand how the "new" environment compares with the "old" environment. Standing up new hardware typically means faster CPU and more RAM. Additionally, new virtualization decisions (i.e. LPARs, VM, etc) may be under consideration as more environments move toward shared and/or cloud infrastructure and how that may impact performance..
A couple of years ago I wrote a comment line about using multiple cells in production. The article covers a common method for providing both high availability and/or continuous availability (i.e. no down time due to planned maintenance) in production environments. However, that isn't the only reason for using multiple cells. My next series of posts looks at using a similar strategy in non-production environments specifically around performance comparison for upgrades to hardware, middleware software or the application. In addition, I'll be investigating the ability to also compare infrastructural changes that may occur due to newer/better hardware capabilities.