Platform9 PCD Upgrade Pre & Post Check Validation

Problem

What are the Pre- and Post-Check Validations for Platform Self-Hosted PCD Upgrades?

Environment

• Platform9 Private Cloud Director- Self-Hosted

Upgrade Preparation

Refer to the following guidelines before starting any upgrades:

• Platform9 Upgrades - Process and Impact
• PCD Release Notes

• PCD Upgrade Documentation

Pre & Post-Check Validations

Control Plane Services Check:

1. Check the current PCD version and Status.

$ airctl status

The airctl status command should display a "ready" status for the deployment status, as well as for region health. Additionally, the desired services must match the ready services for all regions.

$ airctl status

fqdn:                test-pf9.platform9.com

region:              test-pf9

deployment status:   ready

region health:       ✅ Ready

version:              PCD 2025.6-151

-------- region service status ----------

desired services:     29

ready services:       29

​

------------- deployment details ---------------

fqdn:                test-pf9-regionone.platform9.com

region:              test-pf9-regionone

deployment status:   ready

region health:       ✅ Ready

version:              PCD 2025.6-151

-------- region service status ----------

desired services:     77

ready services:       77

• On checking airctl status If task state stuck in **host_upgrade_error**, use the following command to reset it

$ airctl reset-taskstate --region <AFFECTED_REGION_NAME> --region-state ready

2. Check the deployment status. All deployments should be available.

$ kubectl get deployment -n <REGION_NS>

3. Check the hostagent version.

$ NS=<WORKLOAD_REGION_NS>

​

$ kubectl -n $NS exec deploy/resmgr -c resmgr -- curl -s http://localhost:8082/v1/hosts | jq -r '.[] | .host_id + " | " + .status + " | "+ .info.hostname + " | " + .host_agent.version'

Verify that the version of the hostagent is consistent with the current version of the management plane.

4. Check whether the hosts are responding or not from DB.

$ NS=<WORKLOAD_REGION_NS>

​

$ kubectl exec -i -n "$NS" $(kubectl get po -n "$NS" -l du-app=resmgr -o jsonpath="{.items[0].metadata.name}") -c resmgr -- mysql -t resmgr -u root -h $(kubectl get po -n "$NS" -l du-app=mysql -o jsonpath='{.items[0].status.podIP}') -p$(kubectl exec -i -n "$NS" $(kubectl get po -n "$NS" -l du-app=resmgr -o jsonpath="{.items[0].metadata.name}") -c resmgr -- /bin/bash -c "DB=\$(consul-dump-yaml --start-key customers/\$CUSTOMER_ID/regions/\$REGION_ID/db | awk '/dbserver:/ {print \$2}'); consul-dump-yaml --start-key customers/\$CUSTOMER_ID/dbservers/\$DB | awk '/admin_pass:/ {print \$2}'") -e "SELECT id,hostname,responding FROM hosts;

A response of 0 indicates that the node is not connected to the management plane, while a response of 1 indicates that it is connected to the management plane.

OpenStack Services Check:

1. Check hypervisor list

$ openstack --insecure hypervisor list --long

2. Check hypervisor status

$ openstack --insecure hypervisor stats show

3. Check compute service list

$ openstack --insecure compute service list --long

4. Check network agent

$ openstack --insecure network agent list --long

5. Check volume service-list

$ openstack --insecure volume service list --long

6. Filter Inactive VM's

$ openstack server list --all | egrep -v "ACTIVE"

UI Validation

After the upgrade, attempt to access the UI and navigate through multiple pages to verify that the interface is functioning properly. Recommended sample pages include: hosts, cluster, VM, image, and network.

• If these steps prove insufficient or not applicable, reach out to the Platform9 Support Team for additional assistance.