[Linux-cluster] Inconsistency between 'gfs_tool df' and 'df'

[Mark Janssen]

Hello,

One of my GFS1 filesystems was filling up, so I resized the logical
volume and gfs_grow'd the filesystem.

gfs_tool df will correctly show the new maximum and free size, however
normal 'df' still reports the original size, and is reporting a very
full filesystem.

[root at system ~]# gfs_tool df /some/filesystem
/some/filesystem:
  SB lock proto = "lock_dlm"
  SB lock table = "nfsclust:blablabla"
  SB ondisk format = 1309
  SB multihost format = 1401
  Block size = 4096
  Journals = 2
  Resource Groups = 3003
  Mounted lock proto = "lock_dlm"
  Mounted lock table = "nfsclust:blablabla"
  Mounted host data = "jid=1:id=1310722:first=0"
  Journal number = 1
  Lock module flags = 0
  Local flocks = FALSE
  Local caching = FALSE
  Oopses OK = FALSE

  Type           Total          Used           Free           use
%

------------------------------------------------------------------------
  inodes         1739711        1739711        0              100%
  metadata       471881         338342         133539         72%
  data           223141556      193760217      29381339       87%


[root at system ~]# df -h /some/filesystem
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/CluVGgfs-lv--somefilesystem
                      810G  748G   63G  93% /some/filesystem

[root at system ~]# lvs
  LV                 VG       Attr   LSize   Origin Snap%  Move Log Copy
%  Convert
  lv-somefilesystem  CluVGgfs -wi-ao 860.00G


Any idea on how I get '/bin/df' to show the correct 'new' size and
utilization. Currently nagios is putting this filesystem in warning due
to utilization, though it's not full (according to gfs_tool df).

-- 
Mark Janssen  --  maniac(at)maniac.nl  --  pgp: 0x357D2178 |   ,''`.  |
Unix / Linux Open-Source and Internet Consultant @ Snow.nl |  : :' :  |
Maniac.nl      MarkJanssen.nl      NerdNet.nl      Unix.nl |  `. `'   |
Skype: markmjanssen ICQ: 129696007 irc: FooBar on undernet |    `-    |