VMware ESXi and vSphere Cluster Management

VMware ESXi Maintenance Mode Explained

Learn what VMware ESXi maintenance mode does, how to evacuate virtual machines, use it in clusters, compare it with standby mode, and complete a safe maintenance workflow.

What Is ESXi Maintenance Mode?

Maintenance mode is an ESXi host state that removes a host from normal virtual machine workload placement so administrators can safely perform host maintenance. An ESXi host is a physical server running the VMware ESXi hypervisor, while a virtual machine (VM) is a software-defined computer running on that host.

Administrators commonly use maintenance mode before certain software installations, patches, updates, upgrades, hardware work, repairs, and reboots. The host remains managed and available for administrative tasks, but it is no longer intended to run normal production workloads.

What Must Happen to Virtual Machines First?

Before an ESXi host can successfully enter maintenance mode, its running VMs must be evacuated. Evacuation means moving the VMs to another compatible host or powering them off according to an approved maintenance plan.

In a cluster, vMotion can migrate a running VM between compatible hosts with minimal interruption. Distributed Resource Scheduler (DRS) can recommend or automate placement and migration when the cluster is configured to do so. The exact result depends on the DRS automation level, VM rules, available capacity, and migration compatibility.

VM Evacuation Options

Option	Requirements	Downtime impact	Best suited for
Live migration with vMotion	Compatible source and destination hosts, working vMotion networking, accessible storage and networks, and no blocking VM restrictions	Usually minimal interruption	Production VMs that must remain available
Power off the VM	An approved outage window and a recovery plan	Service interruption until the VM is restarted	Environments without vMotion or VMs that cannot be migrated
DRS-assisted evacuation	A configured cluster with suitable DRS automation and enough destination capacity	Usually minimal interruption when vMotion succeeds	Clusters where placement should be automated

Evacuation can fail when destination hosts lack CPU or memory capacity, vMotion is unavailable, storage is inaccessible, networks do not match, CPU compatibility prevents migration, attached devices restrict migration, or host affinity rules prevent placement. An anti-affinity rule can also influence where a VM may be placed, even when it does not directly require a particular host.

Host Behavior During Maintenance Mode

While an ESXi host is in maintenance mode:

New virtual machines cannot be deployed to the host.
Virtual machines cannot be powered on at that host.
Workload placement tools should not return normal production VMs to the host.
Administrators can still manage the host and perform maintenance tasks through vSphere and other approved administrative methods.

These restrictions separate administrative access from workload service. The host may be visible, connected, and manageable, but it is intentionally excluded from normal production VM use until maintenance is complete.

Entering Maintenance Mode in the vSphere Client

Open the vSphere inventory and select the target ESXi host.
Open the host context menu or host actions menu.
Choose Enter Maintenance Mode.
When prompted, select the appropriate VM evacuation option, such as migrating VMs or powering them off according to the available workflow.
Monitor the evacuation tasks and related events.
Confirm that all required VMs have left the host and that the task reports successful completion before starting disruptive work.

The interface may display migration recommendations, warnings, or blockers. Do not assume that selecting the action alone completes the operation. Check task progress, VM power states, host membership, and recent events.

What to Check Before Starting

Check	Why it matters	How to validate	Risk if not addressed
Backups and rollback plan	Provides recovery if an update or hardware operation fails	Confirm recent usable backups, configuration records, and rollback steps	Longer outage or possible data loss
Active workloads	Running VMs must be migrated or powered off	Review the host's VM list and power states	Maintenance mode cannot complete
Cluster capacity	Remaining hosts must support evacuated CPU, memory, and reservations	Review utilization, reservations, admission control expectations, and unavailable hosts	Migration failure or reduced redundancy
vMotion connectivity	Live migration depends on working compatible network paths	Review vMotion configuration, host compatibility, and relevant network status	VMs may require shutdown
Storage availability	Destination hosts need access to required VM storage unless another supported migration method is planned	Verify datastores and paths on destination hosts	Migration cannot complete or VM access may be interrupted
Maintenance scope	Defines expected duration, reboot requirements, and affected services	Review the change plan, vendor guidance, and rollback criteria	Unexpected service impact
Placement rules	Affinity and anti-affinity rules may restrict evacuation	Review VM-host rules and DRS recommendations	Specific VMs remain blocked

Maintenance Mode in a vSphere Cluster

A cluster is a group of ESXi hosts managed together for resource sharing, workload placement, and availability features. Putting one host into maintenance mode normally affects that host, not the entire cluster. Other hosts can continue running workloads if they have enough resources and remain healthy.

Capacity planning is essential. The remaining hosts must accommodate evacuated VMs while preserving the required redundancy and high-availability margin. A cluster that can run normally may not have enough spare capacity to evacuate a host during a period of high utilization or when another host is unavailable.

DRS may select destinations and perform migrations based on its automation level. VM-host affinity rules can require or prefer a VM to run on certain hosts. Anti-affinity rules can keep selected VMs apart. These policies may reduce the number of valid destinations, so review them before maintenance.

Maintenance scheduling should also account for vSphere High Availability expectations, admission control, failure domains, and any requirement to keep multiple copies or tiers of service running. Avoid beginning parallel host maintenance unless the cluster has been designed and approved to tolerate it.

Exiting Maintenance Mode

After maintenance, return the host to workload service by selecting the host and choosing Exit Maintenance Mode from the host actions or context menu.

Confirm the planned update, repair, or upgrade completed successfully.
Validate host health, including hardware status, alarms, boot state, and management connectivity.
Check networking, including management, VM, storage, and vMotion connectivity as applicable.
Confirm required datastores and storage paths are accessible.
Review update status and confirm that any required reboot or remediation step is complete.
Exit maintenance mode.
Verify that workload placement can resume and that DRS or other placement mechanisms behave as expected.
Monitor newly placed or returned workloads and confirm application health.

Exiting maintenance mode makes the host available for workload placement again; it does not by itself guarantee that every VM will immediately return. Placement depends on DRS settings, resource demand, rules, and the chosen operational plan.

Maintenance Mode Versus Standby Mode

Standby mode is a distinct host power state in which an ESXi host is powered down or placed into a low-power condition. Distributed Power Management (DPM) can use standby mode when cluster capacity is temporarily greater than demand and power policies allow a host to be removed from active service.

State	Primary purpose	Host power condition	VM placement behavior	Typical use case	Relevant vSphere feature
Normal operation	Run and manage workloads	Powered on and active	Can host and receive VMs	Routine production service	DRS and other cluster services
Maintenance mode	Safely perform host servicing	Normally powered on and administratively available	New placement and power-on are blocked; VMs are evacuated	Patching, upgrades, repair, or hardware work	vSphere host maintenance workflow
Standby mode	Reduce power consumption when capacity is unnecessary	Powered down or in a low-power state	VMs must be relocated before the host enters standby	Temporary capacity reduction	DPM

Maintenance mode is for safely performing host work. Standby mode is for power optimization. Standby mode is not a substitute for planning VM evacuation before maintenance. If a host must be patched, repaired, or upgraded, use the maintenance workflow and follow the applicable product procedure.

Practical Example: Patching One Host in a DRS-Enabled Cluster

Confirm that the other cluster hosts have enough CPU and memory capacity for the selected host's workloads.
Review DRS automation, vMotion network availability, storage access, VM-host rules, and high-availability requirements.
Select the ESXi host in the vSphere inventory and choose Enter Maintenance Mode.
Allow DRS or vMotion to relocate running VMs.
Monitor tasks and verify that no active workload VMs remain on the host.
Apply the patch and reboot if required.
Validate host health, networking, storage, and update status.
Choose Exit Maintenance Mode and monitor workload placement.

The learning point is that maintenance mode removes the host from workload use while leaving it available for controlled servicing.

Practical Example: No Live Migration Capability

Identify every VM running on the host.
Schedule downtime and gracefully shut down those VMs.
Confirm that the VMs are powered off and no unresolved workload remains.
Enter maintenance mode.
Perform the required maintenance.
Validate the host and exit maintenance mode.
Restart workloads according to the recovery plan and verify application health.

Maintenance mode cannot replace evacuation planning. When vMotion is unavailable, powered-off VMs and an approved outage window may be required.

Operational Workflow

Plan: Define the maintenance scope, duration, risks, rollback method, and service impact.
Protect: Confirm backups or other recovery options and record the current configuration.
Assess: Review active VMs, cluster capacity, storage, networks, DRS, affinity rules, and high-availability requirements.
Evacuate: Migrate compatible VMs or power them off in a controlled manner.
Confirm: Verify that no VMs remain running on the target host.
Service: Perform the patch, installation, upgrade, repair, or reboot.
Validate: Check host health, management connectivity, networking, storage, update status, and relevant alarms.
Return: Exit maintenance mode and monitor workload placement and application health.

Troubleshooting Maintenance Mode Problems

The Host Does Not Enter Maintenance Mode

The usual reason is that one or more VMs remain powered on and cannot be evacuated. Identify the blocking VMs, then check for a valid destination, sufficient CPU and memory capacity, vMotion configuration, storage and network access, CPU compatibility, attached devices, and placement rules. Resolve the migration issue or use a controlled shutdown during an approved window, then retry.

A VM Cannot Migrate

Review vMotion networking and migration compatibility. Confirm that the destination host can access the VM's required storage and networks. Check CPU compatibility, attached devices, VM configuration restrictions, and host-affinity rules. Correct the underlying issue where possible; otherwise, power off the VM only when the maintenance plan permits it.

There Is Not Enough Capacity

Remaining hosts may already be heavily utilized, resource reservations or cluster policies may limit placement, or multiple hosts may be unavailable. Free or add capacity, postpone parallel maintenance, migrate selected workloads elsewhere, or schedule approved VM shutdowns.

Standby Mode Is Being Confused with Maintenance Mode

Standby mode is a power-management state, often associated with DPM. Maintenance mode is a servicing state. Use maintenance mode for patching, repair, upgrades, and hardware work; use standby only when power-management policies determine that host capacity is temporarily unnecessary.

Exam-Relevant Notes

Maintenance mode removes an ESXi host from normal VM workload placement.
Running VMs must be migrated or powered off before evacuation can complete.
vMotion requires compatible hosts and working migration prerequisites.
DRS can assist with placement and migration in a configured cluster.
A host in maintenance mode remains administratively manageable but cannot receive normal VM deployments or power-ons.
Maintenance mode and standby mode have different purposes: servicing versus power optimization.
Cluster capacity, affinity rules, storage access, and high-availability requirements affect whether evacuation succeeds.

For related study, see Maintenance Mode Explained.