Be sure if the server is windows you use ReFS for the storage that will be the backup target. Also would recommend 10GB for network and depending on your storage that hosts the VMs if Fibre Channel then have the VBR server have access to the fabric for faster backups. Also as always ensure lots of CPU and RAM too if an all-in-one device.
@fresfr27 - it sounds like you have more of an “Advanced” Veeam deployment, with separated components not all on your Veeam server..at least your Repository, correct? Is your current Repo physical or virtual? Is the Proxy you’re currently using the ‘default’ on the Veeam server itself, or a separate server(s)? How big is your backup environment?
It sounds like for your Repository, you’re good to go, assuming it’s a physical device. You don’t need all components to be on the Veeam server to do instant restores using Instant Recovery. Is there a reason you want your Veeam server to be a Simple type, or “all-in-one”, deployment? Just start with the Sys Requirements in the Guide and spec out your VBR server using the Sizing Calculator. You can use the Proxy Sizing example in the Best Practice Guide to spec your Proxy . If you don’t have the physical infrastructure to add servers for your Proxy, you can always use VMs and still get pretty decent backup processing performance.
@Chris.Childerhose and @coolsport00 thank you for your advice.
@coolsport00 We have 2 separate Veeam deployments on two isolated domains: one backs up 30 VMs and 30 PCS / Physical servers. The second deployment backs up about 70 VMs and 30 physical server/ PCs.
Each deployment includes a physical tape server, tape library, and a physical Linux server and currently our Proxies and Veeam servers run on VMs.
We have 2 sites and each Veeam deployment backs machines in both sites. The sites are connected with a dark fiber pipe. Our Veeam servers and back up repositories reside in one site and the tape libraries/tape servers reside in the second sites. We have proxies in both sites. One of the domains has Fibre Channel in one site and ISCI in the other. The other domain will have a stretched VXRAILS deployment with VSAN.
Thank you for your help!
Hi @fresfr27 - well, with the amount of VMs you have, you can have your VBR server be a Simple deployment model. I do recommend, at the very least, keeping your Repo on a separate server and direct your VBR configuration backups to it, as well as your backup job files. If you do indeed want to have a cosnolidated server at each site, or one over both sites, just use the resources in the links I provided to spec out your server & you should be good to go.
@coolsport00 Thank you for your reply. I can’t consolidate our two backup instances because one of the domains is isolated because we need to protect the critical VMs it runs.
I am interested in getting very fast restores by adding additional high-speed storage to the new stand along VBR server. I propose making this additional storage a backup target for critical VMS so I can recover them quickly. I will continue to use the existing spinning disk storage for non-critical backups.@coolsport00 is this a good plan?
Many thanks
fr27
Hi @fresfr27 - ok, understood. And, makes sense.
Yes...what you propose to add should be fine.
No problem. Glad to help.
HI Folks,
Can I perform Instant Restores from immutable backup storage with Veeam instant restore? Are there any drawbacks to doing this?
Many thanks
HI Folks,
Can I perform Instant Restores from immutable backup storage with Veeam instant restore? Are there any drawbacks to doing this?
Many thanks
Yes you can just like any other block storage. No issues I am aware of.
Yes, because you’re not changing the backup files; you’re just reading from them. Changes written during the IR process are to redo log files, which then, when you choose to finalize/migrate the IR process, gets merged into the recovered VM disk.
Review the IR process here.