Re: [Libtirpc-devel] svctcp_create with fixed port fails in Linux

[Frank S. <fst...@bi...>]

Hi Steve,

thanks a lot for caring!

Steve Dickson wrote:

> Why not just set the static port through something like /etc/nfs.conf?

The code for the remote uptime server isn't aware of /etc/nfs.conf,
so I can't use that, unfortunately.

>>           /* transp = svctcp_create(RPC_ANYSOCK, 0, 0);  */
>>           /* my code added: */
>>           int sd;
>>           struct sockaddr_in sin;
>>           sd = socket(AF_INET, SOCK_STREAM, 0);
>>           bzero(&sin, sizeof(sin));
>>           sin.sin_family = AF_INET;
>>           sin.sin_addr.s_addr = INADDR_ANY;
>>           sin.sin_port = htons(62222);
>>           bind(sd, &sin, sizeof(sin));
>>           transp = svctcp_create(sd, 0, 0);
>>           /* end of my code added */
>
> Looks reasonable... but where did you come up with  62222?
> Are you sure it is not already being used?

That was just an arbitrary one. I started with 852 which was used in
the past, then thought it might be a problem with ports < 1024 and
used different values above 60000. But it's always the same problem
no matter which port number I try.

> Just curious... is systemd involved with anything?

Yes, the systems are running systemd, and rpcbind is started
via rpcbind.service and rpcbind.socket.

Is it possible that the rpcbind daemon expects sth. different
from the uptime server for opening a fixed port, and refuses
to open the tcp port because my patched code is talking some
"too old" standard? However, there is no indication in the
journal, i.e. no message at all from rpcbind, though, even
not when starting it in debug mode.

> Yeah it is not obvious as to why this is happening... Again I would
> look at systemd "helping out" and make sure the port is not already
> be used.

I'm not sure I understand the comment about systemd: Do you mean it
might somehow prevent opening the port by some security restrictions
or similar? I stopped the rpcbind daemon (including rpcbind.socket)
and started it manually, but still the same error...

-- 
Dipl.-Inform. Frank Steiner   Web:  http://www.bio.ifi.lmu.de/~steiner/
Lehrstuhl f. Bioinformatik    Mail: http://www.bio.ifi.lmu.de/~steiner/m/
LMU, Amalienstr. 17           Phone: +49 89 2180-4049
80333 Muenchen, Germany       Fax:   +49 89 2180-99-4049
* Rekursion kann man erst verstehen, wenn man Rekursion verstanden hat. *