[sheepdog-users] Single disk getting full

Liu Yuan namei.unix at gmail.com
Tue Aug 6 12:23:57 CEST 2013


On Tue, Aug 06, 2013 at 12:04:24PM +0200, Valerio Pachera wrote:
> 2013/8/6 Valerio Pachera <sirio81 at gmail.com>:
> > Hi, on my production cluster (0.6.0_62_gdff7a77) I'm getting some disk full.
> > What do you suggest to do?
> 
> cluster reweight is not going to redistribute data between disks of
> the same md right?

No. Unfortunately, we can't do anything manually to distribute objects. I think
you can simply unplug the disk that is about to full, to prevent the cluster
from stop because of a single disk full.

Thanks
Yuan



More information about the sheepdog-users mailing list