How do I piggyback a virtual machine on a data center server?

In order to optimize the deployment of virtual machines as much as possible, you must carefully analyze the workload to determine which processes and resources they need and when they are needed. For example, if you want to run Windows services on your network, you should be like this:

Domain controllers often require network and processor resources during peak hours, such as in the morning or after lunch.

File and print servers require processor and network resources at off-peak hours, such as around 10 am or around 3 or 4 pm.

Network servers often require network resources. If properly planned, stable traffic should be guaranteed.

SQL Server and Exchange Server require stable disk and processor resources within 24 hours of each day.

Enterprise applications usually have predetermined resource requirements. For example, a corporate payroll application needs to run every two months or every two weeks.

Test and R&D systems are often used during off-hours, or their workload changes are irregular.

Training systems are generally run during the day, but the usual resource requirements are not very large.

Comparison of server resource requirements for different workloads

Not all workloads are busy at any time. In fact, some workloads are often in a "parking" or "no-load" state and rarely run. This requires a reasonable planning - reducing the number of workloads actually running.

Host Server Workload Distribution

Server workloads have different resource requirements at different times, which requires you to properly configure them so that they are not at the same time. The same host competes for the same physical resources -- CPU, RAM, network, or storage resources. Therefore, you should configure as many virtual workloads as possible to avoid overloading the host server with heavy workloads. This means that a single host server can run a data center, a network infrastructure server, a file server, one or more web servers, or even an enterprise application. Of course, the key is to pay enough attention to workloads that have different resource requirements at different times.