Nutanix on VMware: Archived Single-Host Maintenance Sequence
This is an archived Nutanix/vSphere single-host maintenance example, including 2017 shutdown output. Match the actual AOS, ESXi and vCenter release/service procedure before use. The overview lists host maintenance before CVM shutdown, but the detailed source sequence shuts the CVM down before entering ESXi maintenance; do not combine those two orderings without verifying the supported workflow.
1. Verify resiliency before maintenance
Check redundancy/fault tolerance and all relevant health conditions before taking one host out. The source dashboard is visibly an AHV example, despite this VMware article. It shows Data Resiliency OK while overall Health is Critical, 6 hosts/blocks, 227 VMs and HX3310. It illustrates the resiliency panel, not proof of a healthy vSphere cluster. Resolve applicable blockers and verify that the actual cluster can tolerate the planned outage.
2. Move user VMs, leaving the CVM on its associated host
GUI route in the source: vSphere Web Client → select the host → Virtual Machines → select guest VMs excluding the CVM → Migrate → choose the intended compatible destination and complete the wizard. Confirm each migration task, available capacity and any non-migratable workload.
PowerCLI alternative, with interactive credentials rather than a password argument:
Connect-VIServer -Server <VCENTER_ADDRESS> -Credential (Get-Credential)
Get-VMHost -Name "<HOST_NAME_OR_IP>" | Get-VM
Move-VM -VM vm1,vm2 -Destination <DESTINATION_HOST>Replace placeholders with actual identifiers. The inventory command can list the CVM too; do not pass it indiscriminately into Move-VM. The source permits comma-separated guest names. No credential is supplied or stored by this article.
3. Identify the correct Nutanix host entry and CVM
ncli host lsComplete source host-entry example:
Id : 00055b43-73ee-c415-6eb3-000af79c7dc8::48676
Uuid : bf45cebc-ddd7-4f88-985a-6c5c747758ff
Name : C1-VM17.devops.dwzq.com.cn
IPMI Address : 192.168.164.46
Controller VM Address : 192.168.163.146
Hypervisor Address : 192.168.163.46
Host Status : NORMAL
Oplog Disk Size : 420.95 GiB (451,990,218,506 bytes) (1.8%)
Under Maintenance Mode : null (-)
Metadata store status : Metadata store enabled on the node
Node Position : Node physical position can't be displayed for this model. Please refer to Prism UI for this information.
Node Serial (UUID) : J33B83M
Block Serial (Model) : J33B83M (HX5510)The entry links hypervisor 192.168.163.46, CVM 192.168.163.146, IPMI 192.168.164.46 and a host UUID ending 747758ff. This is a Nutanix host record, not a VMware VM UUID. Determine the host identifier accepted by the installed nCLI and confirm the exact host/CVM pairing.
ncli host edit id=<VERIFIED_HOST_ID> enable-maintenance-mode=true
The pictured result is from a different example host/address/UUID than the text entry. It shows the maintenance flag true; do not copy its UUID or infer that every service/evacuation check has passed. Use the version-specific peer-CVM procedure and check the resulting state.
4. Gracefully shut down the target CVM, then maintain and shut down ESXi
nutanix@NTNX-J32X62B-A-CVM:15.2.21.244:~$ cvm_shutdown -P now
2017-06-05 21:37:20 INFO cvm_shutdown:72 Setting up storage traffic forwarding
2017-06-05 21:37:20 INFO cvm_shutdown:75 Establishing Zookeeper connection...
2017-06-05 21:37:20 INFO zookeeper_session.py:76 Using host_port_list: zk1:9876,zk2:9876,zk3:9876
2017-06-05 21:37:20 INFO ipv4config.py:800 Discovered network information: hwaddr 52:54:00:18:84:93, address 15.2.21.244, netmask 255.255.255.0, gateway 15.2.21.254
2017-06-05 21:37:20 INFO cvm_shutdown:94 Executing cmd: sudo shutdown -P now
Broadcast message from nutanix@NTNX-J32X62B-A-CVM
(/dev/pts/0) at 21:37 ...
The system is going down for power off NOW!This is a third archived example (CVM 15.2.21.244, June 5, 2017). Its cvm_shutdown -P now is the supported-script pattern described by the source, not a generic right-click power-off of the storage VM. Confirm the correct target and current vendor procedure before shutdown.
The detailed source then uses vCenter → host → Maintenance Mode → Enter Maintenance Mode, verifies the host state, and uses host → Power → Shut Down. Do not shut down ESXi until guest handling and the CVM transition meet the applicable procedure.
5. Return the host and CVM to service
- After the physical maintenance, power on the ESXi host.
- In the source workflow, exit ESXi maintenance through vCenter.
- Check that its associated CVM starts. The source notes that maintenance-mode shutdown can prevent automatic start; if required, select the correct CVM and use Power On.
- From a healthy peer CVM, remove the maintained host/CVM from Nutanix maintenance using the verified host identifier.
ncli host edit id=<VERIFIED_HOST_ID> enable-maintenance-mode=false6. Verify services, membership and data resiliency
cluster status
cluster status | grep -v UP
nodetool -h 0 ringThe source checks all expected CVMs in cluster-service and ring output. Presence alone is not a complete service-health check; investigate failures or unexpected membership. Wait for the actual cluster’s Data Resiliency to recover to the required green/OK state before restoring workload placement. The stated “few minutes” is not a guaranteed synchronization duration.
Move guest VMs back only as required by the matching version’s placement/DRS policy after recovery is confirmed. The screenshots remain historical examples; they do not establish that maintenance was performed or validated on the reader’s system.