Right now I eventually sorted out much of my issues setting up gluster (thanks to your latest bugfixes it works quite well), now I'm having issues about performance that are more or less due topology. I have one/few main writers and many main readers with a pool of gluster nodes replicating right now the gluster pool resides on a separate switch with the main writer, so I got 1/3 the maximal write speed. I wonder if someone has already a recipe to get the pool replicate by itself and the writer just using one node (in round robin) to speed up things. lu -- Luca Barbato Gentoo Council Member Gentoo/linux Gentoo/PPC http://dev.gentoo.org/~lu_zero
