Hi Another torture test for glusterfs: I run a build on a replicated volue, while killing and restarting server 1 every 45 seconds. After a while, there are directories on the client where some readdir requests get stuck without a reply (but not all: the first one are fine, then one waits the reply forcever). Restarting server 2 fixes the problem. There is nothing relvant in the logs. Known problem? This is on release-3.3 -- Emmanuel Dreyfus manu@xxxxxxxxxx
