Over the years, as a Veeam SE, working with customers to plan their data protection needs, you see a lot of different environments. Some small. Some large. Some mixed; virtual, physical, maybe cloud. But they all want to know the best way to implement Veeam repositories. What is the best practice? You know what comes next. It depends.

So I would often think - how can one utilize the features of Veeam to best suit their own environment. In the larger more enterprise environments I mainly would see servers hosting Windows repos, and later the adoption of Linux. Taking advantage of the ReFS and (my favorite) the XFS file systems. Especially, where SANs were introduced. Targeting 350TB and above. And taking advantage of all the benefits of a modern SAN device. Snapshots, replication, and immutability.

I would see File repositories in smaller environments where SANs were not utilized, or at least not outside of production. This was usually for cost reasons, but I also came to also understand, there are environments that fully embraced virtualization. And introducing something like a physical server to support their backups may be a burden. At the very least not desired. So, how can these customers build an easy to manage solution, without giving up the enterprise features SAN can provide?

Now, I find myself at Pure Storage. And I haven’t stopped thinking on how our customers can solve this problem. What is best practice, or rather what is the best practice to fit any particular customers environment? And the answer can differ from organization to organization. They don’t all have to use block based

So, let’s get into Pure FlashArray File Services. Pure Storage FA File offers the ability to host SMB and NFS storage off the same storage pool as your block storage. So, no data silos. No need to set aside a storage pool just for file services. This allows you to utilize the same platform for all your data needs. And it can provide a simple way to provide repository storage for your Veeam backups, without the need to dedicate extra hardware. And managed through the same interface. Steps how to configure can be found Here

After setup, you will then simply need to configure a Network Attached Storage Veeam Backup Repository. This decouples the repository from a Windows or Linux Server. Allowing the Well, sort of.

When setting up the repository you will be prompted to setup the Share Folder, which will be the share setup on the Pure Array. But you will also notice the choice to select a Gateway Server. This is the server which will write the backup data over the file protocol.

I intentionally left it automatic which makes the choice dynamic. This can work in your environment, but you may need to consider the choice for your environment. You can see from the image below, in a VM with a lot of vmdks, the gateway may differ from the proxy at times, when several proxies are involved in the backup.

Another consideration would be, when dealing with several sites. I would suggest either tying Gateways to sites, or at the very least utilize Proxy Affinity on your repositories.

Then complete the repository configuration with the following settings. Decompress before Writing, and Per-VM Backup Files. This will provide performance and allow the Flash Array to provide higher data reduction, while keeping data reduction across the wire.