Here are a few tips to remember:
* Anything mission-critical needs a twin (applies to servers, network hardware, and even storage).
* Always schedule downtime in your monitoring system.
* Choose your change schedule carefully.
* Devise change plans with multiple input.
* Don’t assume because you can do something then it must be working.
* Get as much input from others as you can to spot any pitfalls.
* Space out changes between redundant systems.
* Use a centralized solution to deploy updates.
* Use auditing and individual accounts.
* Utilize multiple approval methods.
Read Best practices for change management in the data center from Tech Republic