Seems like you are right. I just ran another test and I am no longer seeing these problems. I did make change to config file (added writebehind translator to client) and remounted GlusterFS in between. It had been hours since the files had been copied to GlusterFS before the first test so if it is transient problems some cached info isn't getting flushed. FYI, Larry -----Original Message----- From: Francois Berenger [mailto:berenger at riken.jp] Sent: Monday, December 21, 2009 6:23 PM To: Larry Bates Subject: Re: ls -alR problems on mounted glusterFS I don't know if it is of any help but I have already seen what you describe on NFS mounted volumes. It is like the freshly written files are in a transient state, but after some time all their info is available via ls -l for example.
