DRS not functional/vCLS VM's not deploying post patching ESXI hosts to 8.0 update 3

Products

• VMware vSphere ESXi

Issue/Introduction

After upgrading the ESXi host to version 8.0 U3, the Distributed Resource Scheduler (DRS) is no longer functioning, and the cluster service health status is showing as Degraded/Unhealthy.

Environment

VMware vSphere ESXi 8.0U3

Cause

This issue is primarily attributed to an unsupported or malformed hostname configuration present either within the ESXi host's /etc/hosts file or its general system settings. Common problematic formats that prevent proper hostname resolution and registration include, but are not limited to, trailing characters such as periods (e.g., esxi101., esxi101.example.com.), hyphens (e.g., esx-, esxi101.example.com-), or unintended trailing spaces (e.g., "esxi101.example.com "). Such syntactical inaccuracies prevent the ESXi host and dependent services, notably vCLS, from correctly identifying and utilizing the hostname, leading to operational failures.

Resolution

Procedure for Issue Resolution and vCLS Pod Configuration

The following steps outline the recommended procedure to resolve the identified issue and ensure correct vCLS pod configuration on the affected ESXi host.

1. Enable Cluster Retreat Mode

• Action: Enable the cluster retreat mode in your VMware vSphere environment.
• Reference: Consult VMware Knowledge Base article 316514 for detailed instructions on enabling cluster retreat mode.

2. Correct Hostname Configuration

• Action: Establish an SSH session to the affected ESXi host and execute the following command to correct the hostname configuration.
• Command:
  esxcli system hostname set --host <correct_hostname> --domain <correct_domain.com>
• Note: Replace <correct_hostname> and <correct_domain.com> with the appropriate values for your environment (e.g., --host esx --domain example-domain.com).

3. Verify Hostname Configuration

• Action: Verify the successful application of the hostname change by inspecting the /etc/hosts file. Ensure that the complete hostname, including the domain, is now correctly reflected within the file.

4. Confirm vCLS Pod Setting

• Action: Confirm that the vCLS pod setting is successfully enabled by running the following command.
• Command:
  configstorecli config current get -c esx -g infravisor_pods -k vcls

5. Initiate vCLS Pod Creation

• Action: Initiate the creation of the vCLS pod by executing the following command in the ESXi command-line interface.
• Command:
  configstorecli config current set -c esx -g infravisor_pods -k vcls -p /pod_settings/enabled -v true
• Expected Output:
  cli [root@hostname1:/var/run/crx/infra] configstorecli config current set -c esx -g infravisor_pods -k vcls -p /pod_settings/enabled -v true Set: completed successfully

6. Ascertain vCLS Pod Operational Status

• Action: Ascertain the operational status of the vCLS pods by executing the command below.
• Command:
  inf-cli get pods -n vcls
• Expected Output:
  

7. Restart Essential Services

• Action: Restart the hostd, vpxa, and Infra services on the ESXi host to ensure that all changes are fully applied and take effect.
• Commands:
• /etc/init.d/hostd restart
• /etc/init.d/vpxa restart
• /etc/init.d/infravisor restart

---

Additional Information

vSphere 8.0 Update 3 uses vCLS 2.0 by default. 

NOTE: The hostname can also be changed via DCUI

Additional information on hostname changes: KB307780