thanks for the hint - but we need disabled this option because of a bug with quick-read wich results in "hanging processes".... http://bugs.gluster.com/show_bug.cgi?id=3011 cheers Am 13.07.2011 11:37, schrieb Marco Agostini: > 2011/7/13 Christopher Anderlik<christopher.anderlik at xidras.com>: >> hi list, >> >> we use glusterfs 3.2.1 with this configuration (one server - one client): >> >> Volume Name: office-data >> Type: Replicate >> Status: Started >> Number of Bricks: 1 >> Transport-type: tcp >> Bricks: >> Brick1: gfs-01-01:/GFS/office-data02 >> Options Reconfigured: >> performance.cache-size: 512MB >> performance.quick-read: off >> >> >> we "sometimes" noticed network outages from glusterfs - so for some >> seconds/minutes the client is not able to access the mounted >> glusterfs-share. >> >> > Try without "performance.quick-read: off" parameter, I have not found > it in the documentation: > http://www.gluster.com/community/documentation/index.php/Gluster_3.2:_Tuning_Volume_Options >
