Emmanuel Dreyfus <m...@netbsd.org> wrote: > But further investigation suggests glusterfs may be at fault here, as > lookup still returns the deleted file inode, while readdir shows the > newer one. OTOH I suspect everything would get fixed if the kernel could > send a reclaim for the deleted node. That would cause glusterfs to flush > its stale information.
Update: I beleive I found a fix in glusterfs for the problem. There was some code that checked for inode change and just updated node, instead of returning ESTALE. Changing it to return ESTALE flush the stall information in glusterfs and everything seems fine again. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
