The volume was 2.5TB in size. This is actually a VM running on a Hyper-V cluster.
So I went to the Hyper-V cluster and using failover cluster manager went to settings and found the VHD Set (think .vhds file). Expanded by 500GB to 3TB in size.
Went to SOFS cluster node that currently was the owner of the CSV presented to the OS. Used diskpart to select the disk and partition and attempted extend.
Diskpart showed the correct size. But in disk manager, the volumes listed at the top show the CSVFS drive but it shows 2.5TB in size. But on the lower part where it lists the disks it shows the 3TB and shows it all allocated. In failover cluster manager it shows the wrong size too.
There is an error in the log about a failed CSV snapshot. I didn't think that a shared VHD could have a snapshot taken on it because that it has two different owners (in our case maybe more in other cases).
When we try to copy a file to a location on the drive we see a "File creation error - The requested operation could not be completed due to a file system limitation". This isn't anything special and is just using xcopy.
The volume is a CSVFS volume which is based on NTFS.
It is very strange and maybe this is an issue being caused by SOFS.
This volume has 115 million files 4-8k in size.
Anyone have any thoughts here?
