> We are exploring the possibility of using Gluster in a High Availability > Qmail Cluster. Currently we use NFS to connect 5 qmail nodes to a single > NFS server where the mail is stored. This setup works GREAT for us > without > any problems. (Except that the NFS server clunks out every week or so for > an unknown reason...) > Welcome! sure, it can work fine as NFS replacement. > > We want to move to Gluster and an AFR brick to give us a HA cluster. How > does gluster do with LOTS and LOTS of small files (140GB of files with an > average size of 35k?) My guess is that if it works with NFS it will work > with Gluster, what about when one brick in the AFR goes down? Does > gluster > timeout quickly and resume operations with just one node? > Yes, when a brick goes down, AFR handles it smoothly. User will not see the effect of one machine going offline. -- Amar Tumballi Gluster/GlusterFS Hacker [bulde on #gluster/irc.gnu.org] http://www.zresearch.com - Commoditizing Supercomputing and Superstorage!
