One of the biggest misconceptions about Veeam Backup & Replication, often fueled by competitors, is that it requires the complete server installation in order to run restores. So, this becomes a Single Point of Failure, just like many other solutions from competitors. This is completely untrue: there are two main features in Veeam that make restores possible even without the server installation.
The first one is the catalog: there is no central catalog even if there is a central database. All the informations about the contents of a backup file are saved directly into the backup file itself. So, as long as you have the backup file, you can read informations of its content and start a restore.
The second one is the Extract utility: a little tool, directly available in any Veeam Backup & Replication installation, that can do restores without the complete installation. Let’s see how it works.
First of all, its location. Whenever you install Veeam Backup & Replication, all the binaries of the Extract utility are stored into the main installaton folder
C:\Program Files\Veeam\Backup and Replication\Backup\
There are three files in this folder:
– extract.exe is the windows command line utility
– veeam.backup.extractor is the windows graphical utility
– extract is the linux command line utility
In this way, whatever is your desired way to initiate a restored, you can use one of the three available tools and start a restore. A good plan is to save these files in a separated location, for example the same repository where is save your backups; in this way, whatever happens to your Veeam server, you do not need to do a complete reinstallation of it before being ready to initiate the restore.
When you need to do a restore, on Windows you can use the graphical utility for example. The tool asks you to mount a VBK file, opens it, and shows you all the VMs that are saved into it:
From here, you can select the desired VM and extract it into any folder that can be seen from the Windows machine.
The same activity can be done with the command line, that works in the same way both on Windows and Linux. It can be used interactively (answering to the different questions that the tool asks) or by inserting a command with all the needed parameter. This is the interactive procedure:
As you have seen, the final target is a folder that can be seen from the Windows or LInux machine where you run the tool. This mean you will then need an additional step, using for example VMware Datastore Browser, to move the VM into a vSphere datastore. Since we are copying files that can be pretty large, doubling the steps is not so efficient.
So, here it is an additional “trick”.
Many windows versions have now a client for NFS. It can be easily installed via the Server Manager for example on a Windows 2012 as in this example:
Once you have the client installed, you can connect the Windows machine to an NFS datastore that is already connected to your ESXi servers. For example, I chose this NFS volume in my lab:
Once the datastore is correctly configured to accept connections from the Windows machine where I want to run Extract, I can simply mount it using the Windows mapping utility:
Once you mapped the NFS datastore, you can run as before the Extract utility,and use the mapped drive (Z: in my example) to directly restore the VM into a vSphere datastore in a single operation. from here, you simply have to quickly register the VM into vCenter or the single ESXi and power it on.