Messages in this thread | | | Date | Sat, 22 May 1999 17:09:13 +0200 (CEST) | Subject | Re: NFS hangs under 2.2.6 | From | Trond Myklebust <> |
| |
>>>>> " " == Philip Gladstone <philip@raptor.com> writes:
> I agree. I also think that it is undesirable to have to run > portmapper on a client system. I am quite careful (for security > reasons) not to run chunks of code that act as network servers > (like the portmapper). Anyway, it seems that if you are > prepared to wait out the timeout, then NFS works anyway.
> I'd certainly like to know if there was a mount option so that > I could disable this behaviour.
I must admit, I haven't yet found out why the sunrpc layer insists on using the portmapper. In principle it should be unnecessary if the correct port is passed down from 'mount' to the NFS layer: if the port definition makes it through 'xprt_create_proto' to 'rpc_create_client', portmapper use is disabled for the client.
You definitely will not be able to using locking without the portmapper, so you should always mount with 'nolock' if you want to experiment. The locking code always expects to find the 'rpc.statd' daemon using the portmapper.
Cheers, Trond
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |