How VMware DRS works?
VMware DRS aggregates the computing capacity across a collection of servers and intelligently allocates the available resources among the virtual machines based on predefined rules. When the virtual machine experiences increased load, DRS evaluates its priority.
VMware DRS allows you to control the placement of virtual machines on the hosts within the cluster by using affinity rules. By default, VMware DRS checks every 5mins to see if the cluster's workload is balanced. DRS is needed to be enabled for resource pools to be created.
DRS is invoked by certain actions in the cluster
- adding or removing the ESXi host
- changing resource settings on the VM
Automatic DRS mode determines the best possible distribution of virtual machines and the manual DRS mode provides recommendation for optimal placement of the virtual machines and leaves it the system administrator to decide.
Manual – every time you power on the VM, the cluster prompts you to select the ESXi host where the VM should be hosted. Recommends migration
Partially Automatic – every time you power on the VM, the cluster DRS automatically selects the ESXi host & Recommends migration
Fully Automatic – every time you power on the VM, the cluster DRS automatically selects the ESXi host & migration. Scaled from Conservative to Aggressive
- Apply priority 1 recommendations - affinity rules & host maintenance
- Apply priority 1 & 2 recommendations - promise significant improvement to cluster load balance
- Apply priority 1, 2 & 3 recommendations - promise at-least good improvement to cluster load balance
- Apply priority 1, 2, 3 & 4 recommendations - promise moderate improvement to cluster load balance
- Apply all recommendations - promise even a slight improvement to cluster load balance.
There are three major elements here:
- Migration Threshold
- Target host load standard deviation
- Current host load standard deviation
When you change the “Migration Threshold” the value of the “Target host load standard deviation” will also change. Two host cluster with threshold set to three has a THLSD of 0.2, a three host cluster has a THLSD of 0.163.
While the cluster is imbalanced (Current host load standard deviation > Target host load standard deviation) select a VM to migrate based on specific criteria and simulate a move and re-compute the “Current host load standard deviation” and add to the migration recommendation list. If the cluster is still imbalanced (Current host load standard deviation > Target host load standard deviation) repeat procedure.
How does DRS selects the best VM to move?
For each VM check if a VMotion to each of the hosts which are less utilized than source host would result in a less imbalanced cluster and meets the Cost Benefit and Risk Analysis criteria.