Enable and Configure vSphere DRS on a VMware ESXi Cluster

Learn how to enable vSphere DRS in vCenter, choose automation and migration thresholds, review recommendations, and troubleshoot placement constraints.

vSphere Distributed Resource Scheduler (DRS) is a cluster-level service that helps place and balance virtual machines across ESXi hosts. This lesson explains how to enable DRS through the vSphere Web Client, select an automation policy, review recommendations, and account for migration requirements and placement rules.

What vSphere DRS Does

vSphere Distributed Resource Scheduler (DRS) is a vCenter Server service that makes virtual-machine placement and rebalancing decisions for a cluster. A cluster is a group of ESXi hosts managed together by vCenter Server. An ESXi host is the hypervisor server that runs virtual machines.

DRS evaluates resource demand, available capacity, and configured policies. It considers resources such as CPU and memory when deciding where a VM should run. If resource contention develops, DRS can recommend or perform a vMotion, which is the live migration of a running VM between compatible ESXi hosts.

DRS is not a feature that independently balances VMs on a standalone ESXi host. It requires a vCenter-managed cluster and operates across the hosts in that cluster. A standalone ESXi host can run VMs, but it cannot use cluster-level DRS decisions.

Initial placement: DRS selects an appropriate host when a VM is powered on.
Ongoing balancing: DRS evaluates the cluster after placement and may suggest or perform migrations.
Resource contention: Multiple workloads compete for constrained CPU or memory resources.
Policy awareness: DRS considers reservations, resource pools, affinity rules, compatibility, and available host resources.

Requirements Before Enabling DRS

DRS is coordinated by vCenter Server, so a functioning vCenter environment is required. The hosts must already be members of the target cluster. If you are still building the inventory, review how to add an ESXi host to vCenter Server inventory.

Requirement: vCenter Server management. Why it matters: DRS is configured and coordinated at the cluster level. How to validate: Confirm that the cluster appears in the vCenter inventory and that the vSphere Web Client can manage it.

Requirement: ESXi hosts are cluster members. Why it matters: DRS makes placement decisions among hosts in one cluster. How to validate: Expand the cluster and verify that the intended hosts are listed beneath it.

Requirement: vMotion capability and networking. Why it matters: Ongoing rebalancing commonly uses live VM migration. How to validate: Check VMkernel vMotion configuration, routing or switching, MTU consistency where applicable, and host-to-host reachability.

Requirement: Compatible storage and networks. Why it matters: A destination host must access the VM's required datastores and virtual networks. How to validate: Compare datastore visibility, port groups, VLAN or overlay configuration, and host compatibility.

Requirement: VM migration eligibility. Why it matters: VM hardware, connected devices, reservations, rules, and host compatibility can prevent migration. How to validate: Check migration compatibility and review any vMotion or compatibility warnings.

Requirement: Appropriate licensing. Why it matters: DRS requires a vSphere edition or assigned license that includes the feature. How to validate: Review the license assigned to vCenter and the cluster or hosts.

Before allowing automated moves, review host compatibility, datastore access, network availability, workload reservations, application placement requirements, and any latency-sensitive workloads. A cluster can be technically capable of vMotion while still being unsuitable for unrestricted automated movement.

Navigate to the Cluster DRS Settings

In the vSphere Web Client, select the target cluster in the inventory.
Open the cluster's management or settings area.
Locate vSphere DRS under the cluster configuration pages.
Open the cluster settings editor by selecting Edit.

The exact labels can vary between vSphere releases, but DRS is configured on the cluster object, not on an individual ESXi host.

Enable vSphere DRS

Open the cluster DRS settings editor.
Turn on the vSphere DRS option.
Choose an automation level and migration threshold.
Review the settings and save the updated cluster configuration.

Cluster inventory object > Manage > Settings > vSphere DRS > Edit

Enabling DRS activates the cluster service, but it does not necessarily apply every recommendation immediately. The selected automation level determines whether DRS waits for an administrator, performs only some actions automatically, or performs eligible placement and migration actions automatically.

DRS Automation Modes

The automation level controls two related but different activities: initial VM placement and ongoing rebalancing.

Manual: Initial VM placement: DRS provides placement guidance. Ongoing rebalancing: DRS generates recommendations. Administrator action: Review and apply recommendations manually. Typical use: New or sensitive environments where administrators want maximum control.

Partially automated: Initial VM placement: DRS places a powered-on VM automatically. Ongoing rebalancing: DRS generates migration recommendations. Administrator action: Approve or apply recommended balancing actions. Typical use: Environments that want automatic startup placement but controlled migrations.

Fully automated: Initial VM placement: DRS places powered-on VMs automatically. Ongoing rebalancing: DRS can perform eligible vMotion migrations automatically. Administrator action: Monitor actions and intervene when constraints or workload behavior require it. Typical use: Stable, well-tested clusters with reliable vMotion and consistent capacity.

A conservative starting point is usually appropriate when workload behavior is unknown. Begin with manual or partially automated operation, observe recommendations and migration reasons, and increase automation only after confirming that the cluster, applications, and monitoring processes behave as expected.

Migration Threshold and DRS Aggressiveness

The migration threshold is the DRS aggressiveness setting that controls how readily DRS recommends or performs migrations. Lower sensitivity generally means that DRS acts only when the expected benefit is material. Higher sensitivity allows DRS to respond to smaller imbalances and can produce more activity.

Conservative threshold behavior: Fewer migrations are proposed or performed. Expected frequency: Lower. Balance sensitivity: Responds mainly to significant imbalance. Tradeoff: Less vMotion overhead, but some smaller imbalances remain.

Moderate threshold behavior: DRS balances meaningful differences while filtering minor changes. Expected frequency: Moderate. Balance sensitivity: Practical default for many environments. Tradeoff: Balances responsiveness with migration stability.

Aggressive threshold behavior: DRS is more willing to respond to smaller differences. Expected frequency: Higher. Balance sensitivity: High. Tradeoff: Tighter balance can increase vMotion traffic, transient workload impact, and migration churn.

A tightly balanced cluster is not always a better cluster. Frequent migrations can add overhead, particularly when workload demand changes rapidly. Choose a less aggressive threshold initially, then adjust it using performance history, recommendation frequency, migration duration, and application impact.

Verify DRS and Review Recommendations

Open the cluster Summary view and confirm that DRS is enabled.
Review the displayed cluster balance or overall DRS status.
Open the cluster monitoring area and select vSphere DRS.
Review recommendations, activity, affected VMs, source hosts, destination hosts, and the reasons for proposed actions.

Cluster > Summary                       # Confirm DRS status and balance
Cluster > Monitor > vSphere DRS          # Review recommendations and activity

A DRS recommendation commonly means that DRS proposes moving a VM from one host to another to improve resource balance. In manual mode, the recommendation requires administrator approval. In partially automated mode, initial placement may already be automatic while balancing recommendations remain manual. In fully automated mode, eligible recommendations may be executed automatically.

An imbalance indicator is not necessarily an error. It can simply report that host demand is uneven or that DRS has identified a possible improvement. Treat an alarm, failed migration, compatibility warning, or explicit error as a separate condition requiring investigation.

Example: Reviewing an Imbalance

Suppose one host is heavily utilized while two other hosts have available capacity. DRS may recommend moving a particular VM from the busy host to an eligible destination. In a recommendation-based mode, inspect the proposed source and destination, expected benefit, VM constraints, and application impact before applying it. In fully automated mode, confirm whether the migration was completed by reviewing DRS activity and vMotion tasks.

If no recommendation appears, the cluster may already be sufficiently balanced. Alternatively, DRS may have no feasible destination because of storage visibility, network compatibility, reservations, rules, host compatibility, or insufficient capacity.

DRS Rules and Placement Constraints

An affinity rule is a policy that influences where defined VMs run. An anti-affinity rule requests or requires that defined VMs remain separate. These policies help express application placement requirements, but strict rules reduce the choices available to DRS.

VM-to-VM affinity: Applies to: Selected virtual machines. Placement effect: Requests or requires the VMs to run together. Example: Keep an application VM and a closely coupled service VM on the same host.

VM-to-VM anti-affinity: Applies to: Selected virtual machines. Placement effect: Requests or requires the VMs to run on different hosts. Example: Separate two redundant application components to reduce the effect of one host failure.

VM-to-host affinity: Applies to: Selected VMs and hosts. Placement effect: Prefers or requires VMs to run on specified hosts. Example: Keep a workload on hosts with a required device, license, or application characteristic.

Rules can block an otherwise useful migration. For example, DRS may leave a VM on a heavily loaded host if no destination satisfies its host-affinity rule, datastore access, network requirements, reservation, or hardware compatibility. Document intentional constraints so future administrators understand why DRS does not move a particular VM.

Create and manage these policies in the cluster's VM/Host Rules configuration area. Exact labels vary by vSphere release. Use a dedicated follow-up lesson on DRS affinity and anti-affinity rules when you need to design or maintain placement policies.

Operational Considerations

Maintenance operations: Confirm that workloads can migrate before placing a host into maintenance mode. DRS can assist with evacuation, but it cannot overcome an unavailable destination.
Admission capacity: Leave enough capacity for planned maintenance and host failures. DRS does not replace HA capacity planning; review admission control and admission control policies.
Host failures: DRS and vSphere High Availability have different purposes. DRS balances placement, while HA restarts VMs after qualifying host failures.
Reservations and resource pools: A VM reservation or pool policy can make a destination ineligible even when its average utilization looks low.
Latency-sensitive workloads: Database, real-time, licensing-sensitive, or appliance workloads may require explicit placement review before automated migration.
vMotion dependencies: Validate VMkernel networking, shared or compatible storage, virtual networks, CPU compatibility, and VM device configuration.
Monitoring: Review DRS activity after activation, especially in fully automated mode. Look for repeated moves, failed actions, and migration churn.
Related services: DRS works alongside vMotion, HA, resource pools, and maintenance mode, but it is distinct from each of them.

For storage and host configuration details, review datastores in vSAN where applicable and verify communication between vCenter Server and ESXi.

Troubleshooting DRS

DRS is unavailable or cannot be enabled

Confirm that the selected inventory object is a cluster, not an individual host or VM.
Verify that the environment is managed by vCenter Server.
Check vCenter connectivity and the license assigned to the environment.

DRS is enabled but no recommendations appear

Review the cluster balance; the cluster may already be sufficiently balanced.
Confirm whether the selected automation mode is expected to show pending manual actions.
Examine DRS rules, reservations, host capacity, datastore visibility, network access, and VM migration eligibility.

A recommendation fails or cannot be applied

Validate vMotion VMkernel configuration and host-to-host reachability.
Confirm that the destination can access the required datastore and VM network or port group.
Review CPU compatibility, VM hardware, connected devices, and migration compatibility warnings.
Check whether an affinity rule, reservation, or other resource constraint prevents the move.

DRS performs more migrations than expected

Lower the migration aggressiveness or threshold.
Review performance history and the reasons associated with DRS actions.
Check whether rapidly changing demand, uneven host capacity, or placement rules are causing repeated balancing attempts.

A VM remains on a heavily loaded host

Inspect DRS recommendations and rule compliance.
Verify that a compatible destination has sufficient CPU and memory capacity.
Check storage, network, reservation, hardware, and host-affinity requirements.
If the automation mode is manual, apply a valid pending recommendation after review.

Exam-Relevant Notes

DRS is a vCenter cluster service, not an independent standalone ESXi feature.
DRS uses vMotion for live redistribution of running VMs.
Initial placement occurs when a VM is powered on; ongoing rebalancing occurs afterward.
Manual, partially automated, and fully automated modes differ in who performs placement and balancing actions.
The migration threshold controls how readily DRS recommends or performs migrations; higher sensitivity can increase migration activity.
Affinity and anti-affinity rules constrain DRS placement decisions and can make a migration infeasible.
DRS balances resources; HA provides VM restart protection after qualifying host failures.

Summary

To enable DRS, select a vCenter-managed cluster, open Manage > Settings > vSphere DRS > Edit, enable the feature, select an automation level and migration threshold, and save the configuration. Verify the result in the cluster Summary view and inspect recommendations under Monitor > vSphere DRS. Start conservatively, validate vMotion and destination compatibility, and document placement rules that intentionally limit DRS decisions.