Last year at summit I was discussing gluster with some of the gluster team - and If i remember correctly performance didnt look so hot (with the use of enterprise SAS drives) - I currently work heavily with a proprierty parallel file system which allows us to scale the performance quite quickly and into multple GB.s realm, however this is using fiber storage arrays with dual controllers etc..
If you look at other costly systems such as Islion or Huawei's OceanStor - they tend to sit at the 1GB.s (standard 4 node deployment, with enterprise SAS) - What are your experiences with Gluster and performance ? assuming a 10GB backend
Red Hat
Learning Community
A collaborative learning environment, enabling open source skill development.