Return on premises after a cloud DR

The DR ran, the customer device is up in the Sefthy cloud and the physical server is healthy again. The Console brings the service back home in four steps, and it handles the two longest ones on its own.

Start the return

1. Open the DR page and click Bring device back on-prem, in the phase card.
2. Read the warning and click Confirm.

The Steps to get back on-prem stepper appears with four stages, Perform VM Backup, Perform VM Restore, Log in to the Agent, Completed.

Step 1, the final VM backup

You need a copy of the latest state of the VM, and that is what ends up on the physical machine. If the cloud VM carries the Sefthy Agent, the step runs by itself and you watch the three sub-steps light up in order.

• Credentials delivered to the Sefthy Agent of the cloud VM
• The VM’s Sefthy Agent has logged in
• Final backup completed

The VM stays powered on and the service keeps running, it gets stopped when the restore on the physical machine starts. As soon as the final backup is done the stepper moves on.

If there is no agent on the VM, the Console says so. The Manual procedure (Sefthy Agent and credentials) link opens the box with the agent downloads and the DR credentials, and you launch the backup yourself.

Step 2, the restore on the physical machine

The Console issues the Recovery ISO code and shows the Boot the server from the Recovery ISO and type the code panel. What you do on the server is what you do in a local restore, boot the machine from the USB stick and type the code. The full procedure, error cases included, is in Bare metal restore with the Recovery ISO.

Here, though, you choose nothing. The snapshot is the final backup you just took and the disks are the ones moved to the cloud, so the Console shows the plan already made and only asks for Confirm and start.

From that moment the cloud VM is stopped and the only live copy is the one on the local server.

Steps 3 and 4, closing

Once the restore is over the server boots from the disk. The agent calls the Console, gets fresh credentials and signs in again, so Log in to the Agent closes by itself and the stepper reaches Completed.

On closing, the Console destroys the cloud VM and removes it from the subscription, so it stops weighing on the invoice. The first backup of the server back on site starts, and you get the restore completed email.

If you prefer to move the steps yourself, the Next step button at the bottom of the stepper does the same transitions.

If a step stalls

• Step 1 sits on the first sub-step. The agent on the cloud VM is not reaching the Console. The VM only gets out through the connector, so check that it is powered and connected. Otherwise use the manual procedure.
• There is no agent on the VM. Run the final backup by hand from the manual procedure box, then click Next step.
• The code is gone from the page, the ISO rejects it, or the wrong server connected. The panel gives you New code and Change server / new code, which void the running session and open a clean one.
• The server does not check in within fifteen minutes. The panel warns you. Check that it is powered on and booted from the disk rather than the USB stick, then the cable, DHCP and the network, because on different hardware the NIC MAC address has changed.
• The page looks stuck. The panel refreshes itself. If in doubt, reload the page and look at where the stepper is.