This is current limitation for VxFS due to some immobile structures.

Under heavy FS activity, occasionally, causes some VxFS structural files, mainly inode_list, put at the middle/end of FS, resulting shrinking failure as VxFS can not relocate any structure in the area to be truncated.

The limitation will be enhanced in next version of VxFS.

[How to check whether shrinking FS is possible or not]

First issue 'df -k' command to check the size of the filesystem:
# df -k/test
Filesystem            kbytes    used  avail capacity  Mounted on
/dev/vx/dsk/testdg/testvol
                   5242880   18778 4897603     1%    /test


Next run ncheck utility for theFS.


   # ncheck -F vxfs -oblock=0-14284799/dev/vx/dsk/testdg/testvol

    sectors(18874368)      blocks(9437185)
   -----------------                -----------------
   0/0-0/18874367            0/0-0/9437184

   fileset                fset          match  match      devid/
   name                indx inode indx    inode      sectors                                  name
   ----------            ----  ------  ----    ------      -------------                          ------------------

STRUCTURAL  1    5      1      37    0/14284672-0/14284799

 

fset_1 inode_list has occupied sector 14284672-14284799,which is greater than 6G position, because sector 14284799 is block 7142400(about 7GB where 1k fs block).

Therefore, the FS can only be shrinked up to7GB and not less than 7GB (e.g. 5GB) even FS has only 5GB data or less.