January 27, 2017
Three Ways to Restore Exchange Servers with Datto SIRIS
Backing up Exchange Server is critical for business continuity. People lose emails, drives fail, and servers can be compromised. Below are three efficient ways that Exchange servers can be recovered (running on a Datto SIRIS after a disaster), and then restored back to a production server (returning normal operations). Of course, if you have migrated to Exchange O365, you don’t have to deal with this whole process, because you should have Backupify for O365.
The physical Exchange server’s backplane failed, and it’s going to take three days before new parts arrive.
This is an easy one...locally virtualize and then backup the local virtualization back to the SIRIS.
Unless there is some other terrible stuff on the server.
- Schedule maintenance time for this...not entirely necessary, but there would be a short email outage.
- Find a machine with Outlook that works with your Exchange Server.
- Install Kroll Ontrack Powercontrols onto that workstation (comes free with all SIRIS devices).
- Take one last backup of the VM running on the SIRIS.
- Start a file restore with the SIRIS.
- Power down the VM.
- Power on the server after it is repaired make it live on the network.
- Run Powercontrols on the Outlook machine.
- Select the two Exchange Database (EDB) files (One from the file restore via SAMBA and the other from the old Exchange Server.
- Let Kroll check both for differences.
- Import those differences to the old Exchange EDB.
- Restore Complete.
Virtual exchange server crashes (using VMware).
Virtualize via ESX. This type of restore adds the a restore datastore from the Datto device to the VMware server so you can use the hypervisor’s CPU and RAM.
Even simpler than scenario 1.
- Option 1: Full Storage vMotion back to a datastore of your choice.
- Option 2: Use VMware Converter Standalone to convert a VMDK restore from the Datto device to your hypervisor, in any format you like.
The VMware cluster went down, and there is no VMware High Availability...bigger than JUST a local virtualization. To reduce scope I will just talk about the Exchange Server restore.
In an actual disaster recovery that requires offsite virtualizations, contact support for assistance and Datto will help provide the best and fastest solution to recover and then restore.
- Go to the partner portal and virtualize your Windows Domain server.
Batch virtualize the other VMs as needed.
- Change your MX records as needed
- Have you users connect via VPN, RDP or IPSec
I am only talking about Exchange here because there are many options.
- Have Datto Support send an incremental snapshot down to the Datto device.
- Use Kroll Ontrack (like in the first instance) to compare the current EDB against the old EDB and have Kroll update the old EDB with all the new emails/contacts/calendar info, etc.
This list is by no means the definitive list of possible restores users can perform with a Datto device. In fact, your DRaaS solution should be flexible enough to handle a wide variety of potential disasters. Your solution should be a tool that you can manipulate in any way that you need. If you have other ways you have restored (or failed at restoring) data using any solution, let me know! Message me on Twitter or LinkedIn, I would love to discuss how it worked and if there were more efficient ways of restoring.