Today Veeam had formalized Veeam Backup and Replication 8. This version contains a new interesting component named “Failover Plan”. This feature allows you to schedule properly a failover from the production site to the backup site (disaster recovery plan). This means you can now plan your entire failover in advance, and initiate it with a single click when needed. For example, you can even do a failover plan to schedule a move for the production site.
This guide will show you step by step on how to setup a Failover Plan in Veeam 8. A replication works by taking a snapshot of the virtual machine in VMware then replicating the machine to another host or cluster. In the example, I will show you the operation with a replication on the same node but the operation is similar for another host or a remote host. The goal is to first configure the replication job then the Failover Plan job. We will replicate only one virtual machine.
Setting up a Veeam Replication Job:
To start we need to click on “Replication Job” button in the “Backup & Replication” menu.
Enter a valid name and a description for your replication job and then select “Next”. You have to check DR site boxes if your destination is on a remote site (configure the re-IP etc).
Then, on the right hand side of the next screen select “Add”. You can expand to your virtual center and select to VMs you would like to replicate. If on some of your VMs you do not wish te replicate all of the drives on that server to save space. simply select a server then select exclusions. You can then exclude your drives. If you are unsure what the drives are. In virtual center right click the server, select hard drives and you will be able to see the SCSI ID to tally up what drive is what. You can see I have added only one virtual machine here. Click on “Next”.
The next section is quite important. You get the option to choose your destination host (DR). You can select the resource pool and the datastore for the localisation of your virtual machines. I created a VM folder named “Replica” via the vCenter plateform. Click on “Next”.
On the Job settings screen that follows leave the settings of repository for metadata unless you have a specific respository located in the source site. You get the option to select a repository for the backups metadata. The replica name suffix is what the replicated servers have as a suffix on the destination host (DR). You can let “_replica”. Restore points is as it says, how many replications would you like to be able to go back to. Click on “Advanced”.
In advanced settings, you have to choose the compression level and the deduplication optimization. I recommend you to let default option “Optimal” in compression level. Basically, it’s the best ratio between the compression performance and CPU usage of the proxy server (impact: duration of the replication job). For deduplication, it depends of your environment. In fact, if your destination is a remote site you have to choose “WAN target”. In this lab, I used the same host so I chose “Local target”. Click on “Ok”.
On the Data Transfer screen that follows, implement the settings of source and target proxies. With Veeam B&R version 8, you can use WAN Accelerator technology for replication job (new). So, you have to set the source proxy and the target proxy for the remote site (DR). Reminder, in Veeam software, the Proxy Server is a component which will manage the traffic from a source to a destination. Veeam Backup Server uses the resources of Proxy Server for for transferring (RAM, CPU etc.). For local, check “Direct” option. If you need to set up a off-site replication, you can deploy WAN Accelerator feature (only available in Enterprise Plus licensing) then choose “Through built-in WAN accelerators” option. WAN Accelerator works with cache system, these are folders at source and destination. Click on “Next”.
The next screen is for VSS, uses in your domains admin credentials, this allows veeam to log-in to VSS aware applications such as SQL and Exchange. If you are using any of these applications then enter the information here. Then select “Next”.
Finally, set your scheduling when you would like that the Veeam replication to take place. If you have the network bandwidth there is no reason to not run it continuously, that means in the event of a failure you will have lost the least amount of data. In my case, I run the job periodically every 1 hour. You can also define different ranges (period), click on “Schedule…” button. Click on “Next”.
You can check the summary screen. Checked a tick in replication to start now and the job will run: “Run the job when I click Finish”. If you don’t wait until your schedule begins.
The replication job is starting. You can follow the process and the performances with “Statistics” button in the top of window. The bottleneck is my target (destination host). Basically, the replication job is running on the same node and the physical ressources are limited (8 GB RAM).
As you can see “SRV-WIN” virtual machine is replicating on vCenter interface in my “Replica” folder and the suffix has been added.
Setting up a Failover Plan:
To start the second step we need to click on “Failover Plan” button in the “Backup & Replication” menu.
Enter a valid name and a description for your Failover Plan, you can add PowerShell scripts if you have needs for pre and post failover. Click on “Next”.
In your failover plan, you can add VMs from your replicas and with an easy dialog you can move them up or down in order to set the boot order. For each of them, you can also set a delay so that they don’t start before a previous one has been booted and its services are running. Then, on the right hand side of the next screen select “Add VM”. Click on “Next”.
You can check the summary screen and click on “Finish”.
In “Backup & Replication” menu, a new tab appears: “Failover Plans”. You can see your failover plan inside.
Run a Failover Plan:
To start your Failover Plan, execute a right click on it and choose “Start”.
(If you choose “Start to…” you can choose. the restore point specifying a date.)
The Failover Plan is starting. You can follow the process. Click on “Close” button.
As you can see “SRV-WIN” virtual machine is active in “Replicas” tab (“Backup & Replication” menu). The failover plan controls the replication job.
Since the vCenter server interface, “SRV-WIN_replica” virtual machine is power on.
When your operation is finished, you can failback to your production environment. Execute a right click on “SRV-WIN” virtual machine in “Replicas” tab (“Backup & Replication” menu) and choose “Failback to production…”.
Failack wizard find the replica of your virtual machine. Click on “Next”.
Then, you have to choose the destination for failback operation. 3 possibilities: failback to the original virtual machine, failback to the original the virtual machine but with another location or failback to the specified location if you don’t have original VM. In my case, I chose “Failback to the original VM”, the replica VM will be powered off. Click on “Next”.
You can check the summary screen. Checked the tick: “Power on VM after restoring” and click on “Finish”.
The restore session is starting. You can follow the process log.
VM “SRV-WIN_replica” is shutting down and VM original is now working.
Planned failover is now directly integrated into Veeam B&R 8. This can be used to facilitate data center migrations or to perform maintenance work on your production hosts or production site. This is done without any data loss, but will require a small amount of downtime. When you perform a planned failover, the source VM will be shut down (VM didn’t shutdown in my lab), the last changes will be replicated to the target VM, and the target VM will be started.