TKG cluster resize fails after 1 day of cluster creation
This issue in resizing the TKG cluster occurs because the token to join a cluster has expired. The issue may also be encountered when trying to resize a TKG cluster that was upgraded to VMware Cloud DIrector Container Service Extension 3.1.3 or 3.1.4.
This issue is fixed in VMware Cloud Director Container Service Extension 3.1.3 for newly created clusters.
For clusters created before VMware Cloud Director Container Service Extension 3.1.3, the following workaround is to create a new token and update the RDE:
- Run the following in the control plane node:
kubeadm token create --print-join-command --ttl 0
- In Postman, GET the entity at: https:///cloudapi/1.0.0/entities/. The entity ID can be retrieved from the cluster info page or via
vcd cse cluster info
- Copy the response body in Step 2 and replace the
kubeadm join ... command with the output in Step 1 to form the request body.
- Do a PUT on the same URL as in Step 2 with Content-Type: application/json and with the request body formed in Step 3.
- You can perform the resize operation. If the resize failed, then the operation may be triggered again after Step 4 due to the RDE update triggering the behavior.