Troubleshooting Huawei Cloud Stack Workload Clusters

Start from the preflight error or the stage that ac adm alpha cluster status reports. For failures that every provider shares, see Troubleshooting. Run the ac commands on this page from a machine where ac is logged in to the platform, after ac config use-cluster global.

Creation Stages

SymptomAction
Preflight cannot reach the IAM endpointCheck externalGlobalDomain and schema in the credential Secret, and that the machine that runs ac resolves and reaches iam-apigateway-proxy.<externalGlobalDomain> on TCP/443, or TCP/80 for http. Then run create again.
[1/4] CAPI Cluster stays NotFoundCheck the cluster name: ac get clusters.cluster.x-k8s.io -n cpaas-system.
[2/4] KubeadmControlPlane or [3/4] MachineDeployment does not become readyRead the provider log: ac -n cpaas-system logs deployment/cluster-api-provider-hcs-manager --tail=50. If it names quota or an address that is in use, free it in HCS. Otherwise, contact technical support with the log. A hostname error is set on the owning Machine.
ControlPlaneHAReady is not TrueSee Control Plane HA.
[4/4] ClusterView does not become ready, or the cluster stays Provisioned with nodes NotReadySee Troubleshoot a Workload Cluster Stuck in Provisioned.
Nodes stay NotReady because Kube-OVN is not installed, and the provider logs no sentry errorSee Kube-OVN Is Not Installed.
[2/4] KubeadmControlPlane does not become ready, the instance is ACTIVE, and a pool hostname contains a dotOn the node, run hostname. If it returns the full dotted name, the provider on the global cluster is older than v1.0.1: upgrade it, then see Configure FQDN Hostname on Existing HCS Clusters. If that page's prerequisites are not met, contact technical support.
The Cluster stays in DeletingSee Troubleshoot a Cluster Stuck in Deleting.

Control Plane HA

The Control Plane HA check prints the status, the reason, and a message:

ResultAction
True ControlPlaneHAReadyNone.
False ControlPlaneHAPendingRead the message and wait for the control plane machine or ECS operation it names to finish. If you enabled controlPlaneHA on an existing cluster or changed its policy, see Control Plane HA Placement. If it does not finish, contact technical support.
False ControlPlaneHAFailedRead the message, then check HCS capacity and the policy: with anti-affinity, ECS creation fails when HCS has too few hosts for the rule. If they are correct, contact technical support.

Kube-OVN Is Not Installed

The platform installs Kube-OVN and CoreDNS after the provider hands the cluster over:

ac -n cpaas-system get hcscluster <cluster-name> \
  -o jsonpath='{range .status.conditions[?(@.type=="CoreAddonsHandedOff")]}{.status}{" "}{.reason}{" "}{.message}{"\n"}{end}'
ResultAction
False CoreAddonsHandoffPending or False CoreDNSSkipPhasePendingRead the message and wait for the operation it names to finish. If it does not finish, contact technical support.
False CoreAddonsHandoffBlocked, or True with Kube-OVN still not installedContact technical support with the message.