Network and Storage Protocols
Network and Storage Protocols
Solved! See The Solution
This is embarrassing. The share was pointed at the wrong path. So it was out of space. Too many fingers in the pie I guess.
Please note there is a difference in nature between SMB(CIFS) vs NFS. NFS starts writing immediately to the filesystem and writes as long as there is space. However, CIFS continuously tries to reserve the space before copying (folder/files) to the target. Please check: How much data are you trying to copy over ? How big is the largest folder/file size?
Also, check or share the volume & share-level settings for the SMB(CIFS) at NetApp side?
Related KB:
Robocopy ERROR 112 (0x00000070) There is not enough space on the disk. (microsoft.com)
There are several reasons you might get an "out of space" error.
1. You're actually out of space - this can be at the volume level or the aggregate level (if you're using thin provisioning)
2. You have a quota rule on the volume, user or qtree - this rule would prevent you from exceeding a defined quota.
3. You overran the maxdirsize (memory allocation for files in a single directory)
4. You overran the maxfiles value - each volume allows a set number of files (maxfiles). If you exceed that, you get "out of space." I suspect this is what you're hitting.
Each one of these will log an error in the ONTAP event log. I suggest checking there.
I think I have gone through every one of those documents some twice .
I should have more than enough files, unless I am reading this wrong.
volume show -volume FS01 -fields size, files,files-used vserver volume size files files-used
vserver volume size files files-used
---------------- ------ ------ -------- ----------
Shares_CIFS share01 2.50TB 21251126 97
Filesystem iused ifree %iused Mounted on
/vol/share01/ 97 21251029 0% /share
Real names have been replace with share in most places.
Do you have quotas in place?
What about the event log? See anything there?
::> event log show -messagename *space*
Looks like your copy didn't move any data at all if you only have 97 inodes used (which is the initial default value).
This is embarrassing. The share was pointed at the wrong path. So it was out of space. Too many fingers in the pie I guess.