Restarting an All-in-One Instance in AutoPilot

An all-in-one instance runs every role for your environment on a single server. In the example below, one t4g.xlarge instance carries the jump, storage, and web roles, plus six more. Because a single machine hosts everything, restarting it briefly takes the whole environment offline, so plan the action for a low-traffic window.

Before you start

Confirm you're looking at the right server. On the Servers screen you can verify the instance by its Server ID (for example, i-0859b2e97daf5589d), its private IP (10.10.29.62), and a State of running. A restart only makes sense when the instance is already running.

Steps

1. In the left sidebar, expand Actions and select Servers.
2. Find your instance in the server table and look at the action controls on the right side of its row: start (play), stop (square), and restart (circular arrow).
3. Hover the circular-arrow icon to confirm the Restart tooltip appears, then click it.
4. The instance reboots in place. Its Server ID, instance type, and private IP stay the same, so nothing needs to be reconnected afterward.
5. Click Refresh to update the row and watch the State column return to running.

What to expect

A restart reboots the operating system without recreating the instance, so all attached storage and configuration are preserved. Every hosted role goes down together during the reboot and comes back once the instance finishes booting. Give services a short time to initialize before you rely on the environment again, and use the timestamp beneath the table (for example, "Updated 4 seconds ago") together with Refresh to confirm the current status.

Tips

• If you only need to recover a single service rather than the whole machine, check the Services screen under Actions first, since restarting one service is far less disruptive than rebooting the entire instance.
• Take or verify a recent snapshot under Backups before restarting a production all-in-one instance, just in case.