In an era where you can easily spin up a cloud instance or a new dedicated server, there is an alarming tendency for over-provisioning servers. Fortunately, organizations can combat this trend with regular reviews of server utilization and look at the possibility of consolidation. Here is a step-by-step look at reviewing resources according to Rack Alley:
If you have not done so already, put in place software for centralized monitoring of standard server resources. Disk usage, processor usage, memory usage, and traffic are a few examples of common metrics to monitor.
With the data from the monitoring phase, set thresholds that services should not exceed. For example, you want to ensure that Los Angeles web hosting does not exceed 30% processor usage. Set the thresholds using a calculated maximum in a scenario where it will be on the same server as another service.
Now we look at what services we can consolidate. Often, it helps to consolidate complementary services on the same hardware. For example, there are advantages regarding latency if we combine a web server and the database from which it pulls data.
The consolidation phase is usually the most troublesome, as it requires downtime and several teams to work together in a short time frame. Run trials on separate hardware if possible to avoid any issues.
When consolidation is complete for all possible services, there are several options you can consider going forward. If the organization decides to move to a data center, the process will be faster, easier, and cheaper in the long term.