[Linux-cluster] gfs_grow: Device has grown by less than 100 blocks.... skipping

[Bob Peterson]

On Mon, 2007-12-03 at 20:58 +0100, Leandro Alcalde Perez wrote:
> Hello.
>  
> I need to grow a production GFS, but firts, i have probed with a little gfs.
> The problem is:
> When i try   gfs_grow always get an error:
> [root at localhost ~]# gfs_grow -v /Backup_files
> Device has grown by less than 100 blocks.... skipping
>  
> 
> Even I have created 3 pv, but only the first has dates, the others ones are empty
> [root at localhost ~]# pvs
>  /dev/sdc          VolGroup03 lvm2 a-   1020.00M       0 
>   /dev/sdd          VolGroup03 lvm2 a-   1020.00M 1020.00M
>   /dev/sde          VolGroup03 lvm2 a-     10.00G   10.00G
>  
> And the volum group is:
> [root at localhost ~]# vgs
>   VG         #PV #LV #SN Attr   VSize   VFree 
>    VolGroup03   3   1   0 wz--nc  11.99G 10.99G

Hi Leandro,

You did not include the output from the "lvs" command, which would
show the logical volume information.  I assume from your email that
your GFS file system is located in a logical volume, but that logical
volume needs to have enough free space (> 100 blocks) in order to grow.
So if your LV is inside of VG "VolGroup03", let's assume your LV is
named "leandro_lv".  Then you would need to do something like this:

lvresize -L +10G /dev/VolGroup03/leandro_lv
This extends the logical volume "leandro_lv" so that it can use some of
the free space in the VolGroup03 volume group.  After that,
"gfs2_grow /Backup_files" should work properly.

Regards,

Bob Peterson
Red Hat GFS