veeam replication basics/benefits

[cb122]

I am not from a technical role but for DR and risk purposes I need to gain a basic understanding of the replication features of veeam backup and replication, as I believe it is a product under review by our technical department, to potentially replace our existing backup appliance. I understand the concepts of the backup part of veeam fine e.g. write backups to a 2nd storage device in a different geographical location, but what exactly is the replication part of veeam 'replicating', and how does it integrate/what does it offer above and beyond ESXi, e.g. HA cluster for host failure, is it replicating the next level down, e.g. the virtual servers completely, so that if you lost a VM for whatever reason (as opposed to its host) the 2nd site 'replicated copy' of the VM would kick in? Or am I off track? What kinds of scenarios does the veeam replication save you from the a ESXi HA cluster does not?

Also - as most SAN's already offer a site-to-site replication at HW level, why do you also need a solution that incorporates replication in a backup software solution, aren't they just duplicating the same process of replication of data from one piece of storage to another, or is the software version of replication more reliable than the replication feature that comes with the SAN? There must be a reasoning which I am missing.

[foggy]

Hi, please review this user guide section to get a better idea of what Veeam B&R replication is. Basically, it provides you an exact copy of the VM in the native VMware vSphere format on a spare ESXi host in your DR location, and maintains this copy in sync with the original VM according to the specified schedule, keeping the specified number of restore points to go back in time, if required. This post will likely answer your second question.

[jmmarton]

And to add a little more, SAN-to-SAN replication simply copies the data to the target site. VMware isn't aware of what's there, so in the event of a disaster you have to import the replicated VM into VMware, possibly change IP addresses, and do a little housekeeping to get the virtual machine into a state that it's actually usable. However with our replication everything is ready to be fired up at a moment's notice. There's some additional automation with our "failover plans" that allow an entire group of VMs to be started at once if necessary (e.g. think of the all the VMs needed to run Exchange or SAP). There's even additional automation that can be added to the failover plan via scripting. This is an important distinction between our VM-level replication vs hardware replication at the SAN level.

Joe