[lug] NFS mount woes

[dan radom]

* Jeff Schroeder (jeff at neobox.net) wrote:
> Hey all,
> 
> I have a custom-compiled (think LFS) Linux system I'm using.  I have 
> another Red Hat system serving files via NFS and Samba.  When I try to 
> mount via NFS the custom system, it takes a looooong time (10+ minutes) 
> for the mount to succeed.
> 

rounds like portmap / rpc isn't runnong on the client, or it's blocked

> On the custom (NFS client) system, when I issue
> 
> mount -t nfs redhat:/data /data
> 
> I see the server log file report the mount request instantly:
>  
> May 13 23:47:43 redhat rpc.mountd: authenticated mount request from 
> 192.168.1.111:668 for /data (/data)
> 
> However, the client system chews on it for a while (the mount process 
> can't be killed, BTW) and eventually succeeds.  Then NFS works just 
> dandy.

no.  uninterruptable sleep i/o processes can't be killed

> 
> Does anyone have any insight into this?  Google and HOWTO's don't seem 
> to address a delay that ends up being successful-- only delays that 
> fail.  And yes, portmap *is* running on the client, and /etc/exports is 
> correct on the server.  Other clients (running Red Hat) connect 
> quickly; it's just the custom system.

gotta be portmap / rpc

> 
> TIA,
> Jeff

dan