Hi everyone
I have set up a Hyper-V cluster (Hyper-V server, not Windows+Hyper-V role) using some old hardware and wanted to play around with Storage Spaces. I have simulated a JBOD by presenting individual disks from an old HP MSA via SAS HBAs. I was able to create a new storage pool using the disks, and create a virtual disk + CSV using the pool.
Everything appeared to be functioning OK, but as soon as I move my test VM to a different node than the CSV owner the sequential read/write performance drops from 500MB/s to 100MB/s (around 1Gbit Ethernet speeds).
It would seem that all IO (including by the VM itself) is routed via the CSV owner node (even if the CSV is not in redirected access mode).
Is this by design, or am I missing something?
Thanks
A