DR settings after setup

Changing CPU, RAM and network cards of the recovery machine once the DR is active, and why you should take the agent proposal.

The resources and the network you registered in the wizard decide what the machine that boots in the cloud looks like on restore day. When the server changes, that record has to be updated before, not during the emergency.

Opening the settings

  1. Open the DR from the Disaster Recovery list.
  2. Top right, click the three dots menu, More actions.
  3. Choose Settings.

A line at the top reminds you what you are touching, “CPU, RAM and network interfaces that will be applied to the VM the next time you start a DR”. Nothing you write here touches the protected server.

Resources

The Resources section has two controls, CPU from 1 to 32 cores and RAM from 1 to 128 GB. Drag the slider or type the number in the field on the right. Keep the recovery machine close to the real one, less means a service that comes back slow.

Network cards

Each card is a block with IP Address, Netmask, DNS and VLAN. The Gateway only shows on the first card, which is the primary one. If a connector is attached to the DR, the gateway comes already filled in from the Emergency VPN gateway and you can change it.

The VLAN field takes 1 to 4094 and left empty it counts as 1, meaning no VLAN. Add network card adds one, the bin at the top right of a card removes it. The last remaining card cannot be deleted.

When you confirm, the Console rebuilds the network extensions towards the cloud by itself, one for each distinct VLAN you put on the cards. You do not have to ask support and you do not have to touch the connector. How several cards on several VLANs work is covered in Multiple network cards and VLANs in the DR.

The agent proposal

The Sefthy Agent, the program installed on the server, looks at the network cards on every heartbeat. When it finds a configuration different from the one on record, an amber note shows up on the DR page titled The server network has changed. with a three column table, the field, the Registered value and the one Detected by the agent.

From there you get two buttons.

  • Review and apply opens this same modal with the fields already filled in with the new values. Under every changed field you see the old value struck through, the cards carry a chip saying whether they changed or are new, and the cards the agent no longer sees stay at the bottom as a dashed block. All you do is read and confirm.
  • Reject drops the proposal. It does not come back until the server network changes again.

Going through here beats retyping the values by hand. The data comes from the real server and the proposal closes itself when you save. One thing to watch, if you close the modal without confirming the proposed values are dropped, the amber note stays on the page and you can start over.

The full reading of the seven health indicators is in Server health and network drift.

If a DR starts while you are editing

If someone starts a restore while you have the modal open, the save writes nothing, the modal closes and the Console shows “Settings cannot be changed while a DR is in progress”. Try again once the cloud machine is off. If instead a field fails validation, say a badly written IP or a VLAN out of range, the modal stays open, the field turns red and nothing is saved.

In the press