All Products
Search
Document Center

Restart, reset, and delete a node

Last Updated: Aug 04, 2021

This topic describes how to restart, reset, and delete a node in the Elastic High Performance Computing (E-HPC) console. If errors occur to a node, you can restart or reset it to resolve the errors. If a compute node is no longer needed, you can delete it.

Background information

If you restart, reset, or delete a cluster, you cannot use it as expected for a time. We recommend that you do not restart or reset a node in the Running state unless you need to resolve errors. The following changes happen when you restart, reset, or delete a cluster:

  • After a node is restarted, the jobs that are running on the node are stopped.

  • After a node is reset, the operating system and software return to their initial states.

  • Deleted nodes cannot be recovered. After a pay-as-you-go node is deleted, it is immediately released. If you need to delete a subscription node that has not expired, you must apply for a refund or change its billing method to pay-as-you-go in advance. For more information, see Release an instance.

Restart a node

If errors occur to a node, you can restart it to resolve the errors.

  1. Log on to the E-HPC console.

  2. In the upper-left corner of the top navigation bar, select a region.

  3. In the left-side navigation pane, choose Resource Management > Nodes.

  4. In the Cluster drop-down list, select a cluster.

  5. In the Node Type drop-down list, select a node type.

  6. In the Node List section, select the node that you want to restart. Click Reboot.

  7. In the Reboot Node dialog box, select a restart mode and click OK.

    • Normal Reboot: If you set Reboot Mode to Normal Reboot, restart commands are sent to the node. Then, the operating system automatically terminates all processes and is restarted.

    • Force Reboot: If you set Reboot Mode to Force Reboot, data may be lost. We recommend that you perform a forced restart only when the normal restart mode fails.

Reset a node

If errors persist after you restart the node, you can reset the node to its initial operating system and software.

Note

Only compute nodes can be reset.

  1. Log on to the E-HPC console.

  2. In the left-side navigation pane, choose Resource Management > Nodes.

  3. In the Cluster drop-down list, select a cluster.

  4. In the Node Type drop-down list, select Compute Node.

  5. In the Node List section, select the node.

  6. Choose More Actions > Reset in the Actions column.

  7. In the Reset Node message, click OK.

Delete nodes

You can scale in a cluster by deleting compute nodes.

  1. Log on to the E-HPC console.

  2. In the left-side navigation pane, choose Resource Management > Nodes.

  3. In the Cluster drop-down list, select a cluster.

  4. In the Node Type drop-down list, select Compute Node.

  5. Deletes nodes.

    • To delete a node, select the node in the Node List section, and then choose More Actions > Delete in the Actions column.

    • To delete multiple nodes, select multiple nodes in the Node List section, and then click Batch Delete.

  6. In the Batch Delete message, click OK.