In this new release could we expect to see full Raid 10 speeds from 4 NVME drives in a separate cache pool? Or will the SMB overhead still affect it?
As it stands using the original cache pool design there have been no significant speed differences using Cache: Yes, No or Prefer settings, on 10Gbe, NVME to NVME, Ram Disk to NVME, SSD to NVME transfers. 10Gbe connection verified using iperf3.
If this is as intended and not an expected feature from unraid let me know and I'll move this to a feature request in the correct forum.