On Thu, Aug 30, 2012 at 12:34:35AM -0600, Caleb Call wrote: > Let's bring this back up. According to the document from Citrix, the > ports that are needed are all TCP. Is this correct? Or do we need > UDP for some of these as well (nfs, rpc, etc)? I've got the required > ports opened up (according to the document) and I'm unable to mount my > secondary storage. I'm guessing it's due to needing UDP as well but > would like to confirm that before asking for more ports to be opened > from our neteng team. > > TCP works on port 2049: > > [root@offshore-hv01 ~]# telnet 10.4.43.44 2049 > Trying 10.4.43.44... > Connected to 10.4.43.44. > Escape character is '^]'. > ^] > telnet> quit > Connection closed. > [root@offshore-hv01 ~]# > > However, when I try to mount the nfs share, it just hangs and times-out: > > [root@offshore-hv01 ~]# mount 10.4.43.44:/vol/offshore_vm_images /mnt/nfs > mount: mount to NFS server '10.4.43.44' failed: timed out (retrying). > mount: mount to NFS server '10.4.43.44' failed: timed out (retrying). > mount: mount to NFS server '10.4.43.44' failed: timed out (retrying). > ^C > [root@offshore-hv01 ~]#
<snip> nfs(5) has a section on "Mounting through a firewall". Perhaps that can be of help to debug the problem. You could try forcing the proto=tcp mount option and also check if the portmapper ports are open. tcpdump(1) helps to track it down further. Hth. @shankerbalan -- http://shankerbalan.net/
