On 04/12/2012 11:51 AM, Alex Florescu wrote: > (3) The reason you continue to get I/O errors is probably that the xattrs on > the *parent directory* still indicate pending operations on both sides. You > can verify this with the following command on each brick: > > getfattr -d -e hex -n trusted.glusterfs.dht /a > > > Unfortunately: > getfattr: /a: Input/output error > And when running on any working instance, it says trusted.glusterfs.dht: No > such attribute. Oops, my mistake. It appears that the bricks are at /local on each of the servers, so do the getfattr there. > As per documentation, triggering a self heal is done by > find <gluster-mount> -noleaf -print0 | xargs --null stat (where <gluster-mount> > is /a) - but again, /a cannot be accessed. It still does a stat() on the top-level directory; that's the call that generates the error. Alternatively, to heal just a single item, "ls -ld /a" or "stat /a" would be sufficient.