Open NerdyShawn opened 1 month ago
The cluster itself was still building, then this prevents the terraform destroy
and have to cleanup the resources manually.
Cluster stuck on BUILDING
civo k8s show gpu_operator_civo --region LON1
ID : ca9f0a53-12b8-42da-b3be-3ce049d2faef
Name : gpu_operator_civo
ClusterType : talos
Region : LON1
Nodes : 1
Size : an.g1.l40s.kube.x1
Status : BUILDING
Firewall : gpu_operator_civo-firewall
Version : 1.27.0
API Endpoint : https://74.220.19.126:6443
External IP : 74.220.19.126
DNS A record : ca9f0a53-12b8-42da-b3be-3ce049d2faef.k8s.civo.com
Conditions:
+---------------------------------------+---------+
| Message | Status |
+---------------------------------------+---------+
| Worker nodes from all pools are ready | False |
+---------------------------------------+---------+
| Cluster is on desired version | Unknown |
+---------------------------------------+---------+
| Control Plane is accessible | Unknown |
+---------------------------------------+---------+
Pool (17b25f):
+-----------------------------------------------+----+--------+--------------------+-----------+----------+---------------+
| Name | IP | Status | Size | Cpu Cores | RAM (MB) | SSD disk (GB) |
+-----------------------------------------------+----+--------+--------------------+-----------+----------+---------------+
| gpu-operator-civo-439b-567483-pool-3286-83hg9 | | ACTIVE | an.g1.l40s.kube.x1 | 12 | 131072 | 200 |
+-----------------------------------------------+----+--------+--------------------+-----------+----------+---------------+
Labels:
kubernetes.civo.com/node-pool=17b25fcb-1116-415a-bf25-64b9734073b5
kubernetes.civo.com/node-size=an.g1.l40s.kube.x1
Terraform apply times out when using the default L40s node in LON1. The node seems 🟢 but the v1.27 Talos cluster itself never becomes available enough to pull the kube.config to troubleshoot further.
view of the cluster state
not able to pull the kube.config
The node looks like its ready but the control plane never reaches a ready status.