VMware DRS (Distributed Resource Scheduler), sums up computing capacity across a collection of servers and is allocated to the VM’s based on its needs and requirements. This allows the VM’s to recieve the resources dynamically, when it is required. If in case a VM experiences an increase in load, DRS automatically allocates additional resources, without affecting other VM’s. This is done by continous monitoring of resources by VMware DRS.
Like HA, DRS is also configured cluster wise. DRS allows to prioritize resources to the machine critical applications, that will help aligning the required resources always. VMware Distributed Power Management (DPM) is another feature added with DRS, which needs to be configured explicitly. As the name implies, DPM monitors and manages the power consumption of cluster resources.
How DRS Works?
DRS works based on some user defined rules and policies. This helps in deciding how the resources should be prioritized and shared among VM’s. DRS is configured to work in either of the following three modes:
Fully Automated: DRS will automatically place VM’s onto hosts and will automatically be migrated between hosts to meet resource optimization.
Partially Automated: DRS will automatically place VM’s onto hosts at power on, but vCenter will suggest migration recommendations to meet resource optimization.
Manual: DRS just gives recommendations to help meet resource optimization. All the tasks such as migration should be done manually.
Whenever a VM experiences increased load, DRS first evaluates its priority against the rules and policies. Once DRS is convinced by the resource request from VM, it allocates additional resources. This is done by migrating to a new host(where it meets the requested resources) or by adding more resources to the same host.
It is not mandatory that you should configure DRS in every cluster, but configuring the same can improve resource utilization as well as allocation.