data path during file copy from ESX to a Windows server

[rtheseeker]

What is the data flow path from a source ESX Datastore to a Windows server when using Veeam File copy job. Does the data has to go through the VBR server and then Windows Server or does the data gets directly from ESX Host to the destination Windows server. We have a file copy job place and I never thought about it; however recently we migrated VBR server to a new dedicated ip range for backup (backup VRF) and we ran into this error where the job failed as VBR server could not talk to the destination Windows server on port 2500.

We will be slowly migrating many jobs into Veeam server and this file copy job copies TB of data; so just want to make sure we do not want to runinto any issues due to this copy job and may need to plan for alternatives

[david.domask]

Hi Rajeev,

Correct, the Veeam server will make the connection(s) and handle the copying in this case.

What are you migrating with the File Copy job if you don't mind sharing? Are you moving backup files by chance since it's TiB of data? While the File Copy job is great for simple copies of like configuration files, there may be more performant ways depending on what you're migrating from/to and the limitations in place.

[rtheseeker]

So we have a horizon VDI solution in place, it has lots of VMDK files as templates which cant be attached to any VM hence cant be backed up using the VM level backup; so the current solution is that we copy these files to a Windows server and then we do a backup of the Windows server. The restore method is resotring those files from backup repository back to the Windows server and then doing a file copy again back to the datastore; this solution was proposes to handle any crypto event as the backup storage is immutable

[david.domask]

Got it thank you for the information; is it possible to expose the same directory as an SMB or NFS share perhaps or just add the same server as a File Server for backup with our Unstructured Data Backup? Or even Veeam Agent for Windows?

With adding it as a Windows File Server OR an SMB 3.0 share that supports remote VSS, the data will be ready from a VSS snapshot instead of production and you'll be able to backup directly from the main server and save the copy operations. I'm not as familiar with VDI, but as I get it there should be a persistent storage for the disks that get attached to the clients, right? Is it some VM in the VMware environment that could be protected?

Basically it kind of sounds like your concern is the snapshot, but we have a lot of solutions for protecting this data without the Vmware snapshot, and just seems easier at first blush, but maybe I'm missing elements where that's not feasible.