Author: Jason Mattox, VP of Support and Product Management, Vizioncore
With the release of vRanger Pro 4.0 we are introducing a more efficient and scalable way to perform backups. So on that note, let’s take a second and reflect on how many VMs you have. Think back to before you had virtualization, would you have as many VMs as physical machines if you didn’t have virtualization? In most cases the answer might be no. But, because it’s as easy as a “right click” to deploy and sysprep a VM, most likely your VM count is a lot higher than if you had all physical hardware. This is a leading cause for “VM Sprawl”, “Backup Sprawl” and even possibly “Backup Window Sprawl”, as the number of VMs you want protected at the image level has grown exponentially. As your VM count grows you want to keep leveraging your investment in virtualization and the VM portability or DR/BC that comes along with it.
One of our many goals of the vRanger Pro 4.0 DPP release is to help reduce the overall backup window. With the introduction of our "Direct to Target Architecture", we are proving a scalable way to backup a large amount of VMs in short period of time. If you look at proxy based solutions you have to scale out your proxy servers along with your storage, this can be costly and complex to manage. With our direct to target architecture you only need to add more storage targets for the more backups you want to do at one time. This type of backup traffic spread across many ESX hosts can send more data than a single storage target can handle. If your storage target can only handle let’s say 20 backups at one time but you want to be able to do 40, you can just configure another storage target in vRanger Pro 4.0 and you would be able to scale horizontally.
This type of architecture puts the vRanger server out of band to the backup traffic instead of in band like proxy based solutions. This allows you to run vRanger Pro 4.0 DPP on a VM or physical machine, this no longer matters. If you were to put a proxy based solution in a VM you would end up using 2X more virtual resources to get the backups from the ESX host to the target storage because the data would have to go from the ESX host to proxy VM, then to the storage target.
Another great benefit of this architecture is the centralized management of your backups over many datacenters. For example, if you have datacenter A and datacenter B each with their own storage you could create backup jobs from a single install of vRanger Pro 4.0 for both datacenters. When you’re creating the backup job you would just pick VMs at datacenter B and your storage target at datacenter B, this will force the backup traffic to stay between the ESX hosts and the storage target at datacenter B.
The current direct to target architecture protocols that will be supported in 4.0 will be CIFS (windows share or Unix emulating window share) and SFTP. The SFTP direct to target protocol is good for backing up from the DMZ and only needing to open port 22 from the DMZ to the LAN. When using SFTP you still get all the great features over this secured port like full, incremental, differential and instant file level restore.
Thanks for reading, as I just wanted to share a major architectural shift from vRanger Pro 3.x to vRanger Pro 4.x Data Protection Platform (DPP).
Learn more about vRanger Pro 4.0 DPP at http://www.vizioncore.com/evolution.