Removing a Node
Scenarios
- If there are too many idle nodes in the cluster, you can remove some nodes to save money.
- If a faulty node in the cluster cannot be manually restored, you can remove it and then add a new node to the cluster as needed.
Precautions
- Master nodes cannot be removed.
- A compute node can be removed only when its cluster is in the Running or Abnormal state.
- A compute node can be removed only when other nodes in the cluster are in the Running state.
- Before removing a compute node, you need to ensure that no jobs are running on the node. Otherwise, the running jobs will be abnormal.
- Attached data disks will be retained and continue to be billed after the node is removed. To stop billing for these data disks, delete them manually.
Removing a Single Node
- Log in to the HPC-S² console.
Removing Nodes in Batches
- Log in to the HPC-S² console.
- In the navigation pane on the left, choose Nodes.
- Select the target cluster from the cluster drop-down list.
Select the nodes to be removed and click Remove on the left of the cluster name.
- Pay-per-use
If you select Delete cloud server, the cloud servers will be deleted. Otherwise, the cloud servers will be retained. You can view the cloud servers on the corresponding service console.
- Yearly/Monthly
You need to manually unsubscribe from yearly/monthly cloud servers.
- Yearly/Monthly and Pay-per-use
If you select Delete cloud server, pay-per-use cloud servers will be deleted and yearly/monthly cloud servers will be retained and continue to be billed. To stop billing for these yearly/monthly cloud servers, unsubscribe from them manually. If you do not select Delete cloud server, all cloud servers will be retained.
- Pay-per-use
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot