Hi,
Our companies clusters run Ontap 9.6sp4 with many CIFS SVMs and contain a few hundreds of ntfs volumes that are quite large 5 to 10 TB and even more with lots of Directories, Sub-Directories and files.
As mentionned, some volumes contain lots of child directories with lots of files. When CIFS administrators right click a folder in the windows explorer.exe to have look at the current Directory size, the count and sizing stops at some point ... and returns incorrect values f.ex. 13GB with 2300 files etc ... in reality there are 1.6TB with a few 100 of thousands of files in that directory:
volume\Directory\SubDirectory\SubDirectory\1000 of files\SubDirectory\SubDirectory\SubDirectory\1000 of files
volume\Directory\SubDirectory\SubDirectory\1000 of files
volume\Directory1\SubDirectory\SubDirectory\SubDirectory\1000 of files
etc ....
total 1.6TB
windows explorer (right click properties on folder) +/- 15GB
I guess there is probably some kind a CIFS threshold (client or SVM side) that avoids excessive CIFS ressource consumption for windows clients.
The only tool that we were able to find is the Netapp xcp.exe tool. The tool return the correct size, but we would really appreciate to now if there are other possibilities.
f.ex:
xcp scan -stats volume\Directory
Gives the correct size 1.6TB
The problem doesn't happen on volumes containing a limited amount of child directory objects, f.ex. we are able to right click a folder with 2 level of subdirectories which contains almost 10TB of video files.
volume\Directory1\empty
volume\Directory2\30-40000 movie files
volume\Directory3\30-40000 audio files
total 9tb (is also showed correctly in windows explorer.exe)
Has anyone an idea what could cause this type of issues in Windows or Netapp?
How can you get the correct directory size of any random "directory" (not a volume size!!!) in Ontap (ideally in the "Netapp Powershell toolkit"?
Thank you very much for any kind of help!
Regards,
Didier
