Why do pods remain in pending state?

Virtual Private Cloud Classic infrastructure

When you run oc get pods, you can see pods that remain in a Pending state.

If you just created the Red Hat OpenShift cluster, the worker nodes might still be configuring.

If this cluster is an existing one:

• You might not have enough capacity in your cluster to deploy the pod.
• The pod might have exceeded a resource request or limit.

This task requires the IBM Cloud IAM Administrator platform access role for the cluster and the Manager service access role for all namespaces.

If you just created the Red Hat OpenShift cluster, run the following command and wait for the worker nodes to initialize.

oc get nodes

If this cluster is an existing one, check your cluster capacity.

1. From the Red Hat OpenShift clusters console, select your cluster.

2. Click Red Hat OpenShift web console.

3. Check if you have enough capacity in your cluster to deploy your pod.

4. If you don't have enough capacity in your cluster, resize your worker pool to add more nodes.

   1. Review the current sizes and flavors of your worker pools to decide which one to resize.

      ibmcloud oc worker-pool ls
      

   2. Resize your worker pools to add more nodes to each zone that the pool spans.

      ibmcloud oc worker-pool resize --worker-pool <worker_pool> --cluster <cluster_name_or_ID> --size-per-zone <workers_per_zone>
      

5. Optional: Check your pod resource requests.

   1. Confirm that the resources.requests values are not larger than the worker node's capacity. For example, if the pod request cpu: 4000m, or 4 cores, but the worker node size is only 2 cores, the pod can't be deployed.

      oc get pod <pod_name> -o yaml
      

   2. If the request exceeds the available capacity, add a worker pool with worker nodes that can fulfill the request. For more information, see Adding worker nodes to Classic clusters or Adding worker nodes to VPC clusters.

6. If your pods still stay in a pending state after the worker node is fully deployed, review the Kubernetes documentation to further troubleshoot the pending state of your pod.