I have recently been playing with a pair of microservers, and have put Server Core 2016 on them + Starwind Virtual SAN giving me a nice Hyper-V failover cluster. If testing this turns out well, I'd like to use it for both the home lab and normal home server. The issue is all of our films, TV series, files etc come in at around 8TB, which wont be part of the virtual SAN. This means that 4x 3TB disks I currently have will be distributed between the microservers. I'd like to continue to use Drivepool, and it looks like the best way to achieve what I want is to share out each individual hard disk, and add the lot into CloudDrive running on a highly available VM running Drivepool. I don't think I will want to use any features in CloudDrive such as pre-fetching or local caching; I only want the shares presented as physical disks for Drivepool.
What are everyone's thoughts on this? When one of the microservers is shutdown, the VM running Drivepool should failover and continue to provide a service. I take it CloudDrive will gracefully handle the shares going missing from the shutdown server?
stuart475898 got a reaction from Antoineki in Using CloudDrive + Drivepool for distributed filesystem