Hi, We've been running glusterfs client and server in a single process in our single backend cluster for a few months. The read/write speed on the backend is significantly faster (around 10 times faster) than client accessing mount over TCP. Unfortunately, running client and server in one process expos coherence issue as described in gluster bug 926. We split up client and server as of last Thursday. Tejas - do you have a rough idea on when 926 will get fixed? Thanks, Lei
