Cloud Connect Replication Partial Failover – Example
Veeam Cloud Connect Replication has been part of Veeam’s Backup & Replication product since version 9 was released in early 2016 and like Cloud Connect Backup before it, Replication is starting to get traction in the market due to it’s ease of use, intuitive interface and best of breed disaster recovery networking technologies that are all baked into the core Backup & Replication product.
Without doubt one of the best/little known features of VCCR is the partial failover. Most disaster recovery scenarios focus on the total failure of all systems due to natural events or localised datacenter issues however the most common outage occurs at an virtual machine or instance level…this is generally an issue with the application or the operating system. With that, the ability to failover a single VM at a time is an often overlooked feature when looking into replication and disaster recovery platforms.
Veeam provides partial failovers within Cloud Connect Replication and once a VM has a replica copied up to the cloud provider you have the ability to perform a partial failover. I’ve created a video showing the process involved to initiate a partial failover which starts the VM replica up in the cloud providers hosting platform and then creates a L2 Tunnel via Network Extension Appliances that are deployed at the production and cloud ends. For an explanation of the Network Extension Appliance click here. Without diving into the specifics of what’s happening underneath the surface the NEAs talk to each other via the Cloud Connect Gateway and bridge the layer 2 network providing layer 3 access so that the replica VM that’s been partially failed over can communicate with VMs on the production network and vice versa.
This effectively means services and applications will be available over the internet without the need to employ expensive WAN connectivity mechanisms…the NEAs do all the work for you. In the demo video I am simulating the failure of a VM that hosts a WordPress site. That VM is brought up at the cloud providers end and, as can be seen in the video is shown to be running in a failed over state. From there I trigger a failback which replicates any changes made at the cloud end during the failover state back to the production site on premises. Once I am happy I commit the failback and the VM resumes normal operations on-premises.
The uptake of Cloud Connect Replication through Veeam’s VCSPs has been steady and we are seeing the number of replication VM licenses reported gather pace and grow month on month. As the Backup and Disaster Recovery markets mature I fully expect Cloud Connect to be a central part of our customers 3-2-1 rule of backup and availability with Cloud Connect Replication becoming the best of breed Replication/Disaster Recovery as a Service offering in the market.