Rebooting a node restarts its operating system in place. Its plan, storage, IPs, and configuration stay exactly as they were.
Open the node’s Actions menu and click Reboot; it is available only while the node is Running. The dialog warns you: an active disk-heavy process, like a database or a backup, could suffer data corruption or loss during the restart. If you are not sure, wait for one to finish before you confirm. Once you confirm, the node restarts and returns to Running right away, without showing any other status in between.
Reboot when the node needs an ordinary restart, the same as restarting any computer. When a restart is not enough, a fresh operating system is the next step. See Reinstall a node: it erases everything on the node to get there.
When the node reads Running again, connect and check that your services started. For other reasons Reboot can be greyed out, see Troubleshooting node actions and lifecycle. To reboot a node through the API instead of the portal, see the node-action operation in the API reference.