Hey everyone! I'm currently working on a Horizon 8 setup for a client who needs a disaster recovery (DR) solution that spans two data centers. Here's the deal: Site A is our working production site, while Site B is still being deployed as a DR site. We're using Instant Clone pools with FSLogix at Site A, and the FSLogix profiles are stored in containers. Everything's running smoothly with no user issues so far.
Now, as we're getting Site B ready just in case Site A goes down, I'm grappling with how to manage FSLogix container availability across both sites. Users connect via LAN from their laptops, and I'd like to clarify that no UAG is involved for these users—they're just using Horizon.
We're considering two potential designs. The first option is to have one SOFS cluster that stretches over both sites. This will keep everything in sync, but I'm worried about ensuring users only connect to their respective site's SOFS nodes. The second option involves setting up two independent SOFS clusters, one for each site, with separate UNC paths for each. This will keep things separate, but it raises a bunch of concerns around syncing the containers without risking data corruption.
I'm reaching out to see if anyone here can share insights on the most reliable way to make FSLogix profiles available in a disaster recovery setup, while avoiding data loss and maintaining performance. Any real-world experiences or advice would be greatly appreciated!
1 Answer
To start off, make sure you have good connectivity between your sites—this is crucial. How much storage do you have on both sides? You'll need enough to cover all user profiles. Also, how many users are we talking about? If bandwidth is a concern, I would suggest planning ahead to get that upgrade to 10Gbps. It’ll make a huge difference!
Thanks for the tips! Storage isn't an issue, but yes, bandwidth could be a hurdle right now. They plan to upgrade to 10Gbps, but it might take some time. We’ll keep that in mind.