Am 16.06.2011 07:55, schrieb Christopher Anderlik: Hi! > After upgrading to 3.2.1. and turning quick-read off > (http://bugs.gluster.com/show_bug.cgi?id=3011) glusterfs works very fine > at the moment :-) Wow... that was helpful! After having recreated all volumes and selecting performance.quick-read off (*) performance.stat-prefetch on (*) performance.io-thread-count 64 performance.io-cache on (*) performance.cache-size 64MB performance.write-behind-window-size 2097152 where especially (*) were modified, I have a huge positive impact. Just rsync'ing the whole mail partition of 12 GB, the serverload remains below 2 with rare exceptions. Before with the old settings, server load raised up to five. Also the sync process runs significantly faster. Regards,.
