HA! that was it. dolt! thank you. i was going crazy looking at other stuff. -matt ------- Matt Hodson Scientific Customer Support, Geospiza (206) 633-4403, Ext. 111 http://www.geospiza.com On Nov 3, 2010, at 11:26 AM, Vikas Gorur wrote: > > On Nov 3, 2010, at 11:18 AM, Matt Hodson wrote: > >> I just installed distributed gluster FS on 2 CentOS 5 boxes. >> install and configuration seemed to go fine. gluterd is running. >> firewalls/iptables are off. however for the life of me i cannot >> nfs mount the main gluster server from either a OSX or a CentOS 5 >> box. I use NFS often and have a fair amount of experience with it >> so i've reviewed most of the common pitfalls. >> >> here's the command that fails from centos: >> $ sudo mount -v -t nfs 172.16.1.76:/gs-test /mnt/gluster/ >> mount: trying 172.16.1.76 prog 100003 vers 3 prot tcp port 2049 >> mount: trying 172.16.1.76 prog 100005 vers 3 prot udp port 909 >> mount: 172.16.1.76:/gs-test failed, reason given by server: No such >> file or directory >> >> and the same one from OSX 10.5 >> sudo mount -v -t nfs 172.16.1.76:/gs-test /gluster/ >> mount_nfs: can't access /gs-test: No such file or directory >> >> what's weird is that i can mount actual dirs on the gluster server, >> just not the gluster VOLNMAE. in other words, this command works >> fine because it's mounting an actual dir. >> $sudo mount -v -t nfs 172.16.1.76:/ /mnt/gluster/ > > You have the kernel NFS service running. That is why you can mount > regular directories on the gluster server. > > When you try to mount Gluster the kernel NFS server is actually > looking for a directory called /gs-test, which ofcourse does not > exist. You need to stop the kernel NFS service and stop and start > the gluster volume. > > ------------------------------ > Vikas Gorur > Engineer - Gluster, Inc. > ------------------------------ > > > > > > > >