Re: Need help for file sharing -- linux --> linux boxen
Greetings Nate,
Before I begin, a big thank you for trying to help me with this!
On 04-05-03, nate transmitted the following;
% C-Cose Masters said:
%
% > pmap_getmaps rpc problem: RPC: Unable to receive; errno = Broken pipe
% > Stopping portmap daemon: portmap.
% > Starting portmap daemon: portmap.
%
% next thing would be stop portmapper
%
% /etc/init.d/portmap stop
%
% and run netstat to be sure nothing is using port 111
%
% netstat -an | grep 111
%
% also be sure portmapper is really stopped by checking the process
% list. if something is using port 111 use fuser to determine what
% it is:
%
% fuser -n tcp 111
% fuser -n udp 111
%
All of the above done; the net result was that portsentry was still in
use and listening on that port, so I killed portsentry.
% if both of the above tests return nothing my next idea would be to
% run strace on portmapper as it starts and log the output:
%
% strace /sbin/portmap >&/tmp/output.log
%
Here's the output of that file ...
## begin paste portmap strace ##
execve("/sbin/portmap", ["/sbin/portmap"], [/* 22 vars */]) = 0
uname({sys="Linux", node="grommit.mastersdaley.sys", ...}) = 0
brk(0) = 0x804bb1c
open("/etc/ld.so.preload", O_RDONLY) = -1 ENOENT (No such file or directory)
open("/etc/ld.so.cache", O_RDONLY) = 3
fstat64(3, 0xbfffee10) = -1 ENOSYS (Function not implemented)
fstat(3, {st_mode=S_IFREG|0644, st_size=36098, ...}) = 0
old_mmap(NULL, 36098, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40012000
close(3) = 0
open("/lib/libwrap.so.0", O_RDONLY) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0@ \0\000"..., 1024) = 1024
fstat(3, {st_mode=S_IFREG|0644, st_size=24328, ...}) = 0
old_mmap(NULL, 29092, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4001b000
mprotect(0x40021000, 4516, PROT_NONE) = 0
old_mmap(0x40021000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x5000) = 0x40021000
old_mmap(0x40022000, 420, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40022000
close(3) = 0
open("/lib/libnsl.so.1", O_RDONLY) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0d;\0\000"..., 1024) = 1024
fstat(3, {st_mode=S_IFREG|0644, st_size=69132, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40023000
old_mmap(NULL, 76448, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40024000
mprotect(0x40034000, 10912, PROT_NONE) = 0
## end paste ##
The only line that means anything to me is the second, identifying my
hostname, which is correct given the value I have in the various
"hostname" files. In the very beginning (pre-slink) my system complained
at various intervals about not having a FQDN so I use the *.sys for this
desktop's hostname. Don't know if that's a clue or not.
After portmap was restarted as above, I attempted ~another~ re-install
of nfs-user-server and nfs-common, with the same results as always:
Reading Package Lists... Done
Building Dependency Tree... Done
The following NEW packages will be installed:
nfs-common nfs-user-server
0 packages upgraded, 2 newly installed, 0 to remove and 4 not
upgraded.
Need to get 0B/142kB of archives. After unpacking 463kB will be
used.
Pre-configuring packages ...
Selecting previously deselected package nfs-common.
(Reading database ... 109839 files and directories currently
installed.)
Unpacking nfs-common (from .../nfs-common_1%3a1.0-2_i386.deb) ...
Selecting previously deselected package nfs-user-server.
Unpacking nfs-user-server (from
.../nfs-user-server_2.2beta47-12_i386.deb) ...
Setting up nfs-common (1.0-2) ...
Starting NFS common utilities: statd lockd.
Setting up nfs-user-server (2.2beta47-12) ...
Starting NFS servers:Cannot register service: RPC: Unable to
receive; errno = Connection refused
dpkg: error processing nfs-user-server (--configure):
subprocess post-installation script returned error exit status 1
Errors were encountered while processing:
nfs-user-server
% it may give some clues..though strace is REAAAAAAAAAALY verbose,
% finding the problem is usually not too easy for me at least,
% I'm not a developer, but sometimes it can be found... try posting
% the log file on a web/ftp site if you can't decipher it.
%
% and after that I'm not sure what to suggest, are you running
% woody? if not I suppose it could be a bug..
%
Yep, running a 99.9% woody (3.0r1) system, upgraded from potato; with
Adrian bunk back-ports for X & OpenOffice (for woody); pan, glabel, and
other small pkgs (mostly image manipulation) from testing; amended
kernel-2.2.22 to enable fb support. Nothing in my package list jumps out
at me, except that libc6 is likely from testing.
I recall, when googling for this that of the 2 hits I had for the RPC
connect error, that both posts mentioned former issues with libc, and
the incorrect return of a "connect error" message when communicating via
tcp to rpc services (that post listed the exact output of my rpcinfo
results). But those posts all referred to libc5.
I'm now at that point where I'm ready to forget nfs file/directory
sharing and fall back on ftp between the two Linux boxen, but I'd rather
use the right tool for the right job!
The truly confusing thing is that when I remove nfs-*-server, the
following happens:
Stopping NFS servers: mountd nfsd.
^^^^
It appears that nfs-user-server is installed sufficiently to provide the
nfs daemon that I want, but barfs at the stage of recognizing /
implementing RPC services.
I have to wonder, am I over-thinking this whole process? Is there another
convenient way to allow me to view my files (/home dir) on both machines
from the other (desktop and laptop) without resorting to NFS? The only
convenient method I can think of is ftp. I know that I could telnet
(over secure line) or ssh into one machine from the other should I need
to do admin tasks remotely, but that also seems "ungood" (TM 1984) for
my purposes.
In any case, you continue to be very helpful, for which I am thankful.
% nate
%
C-Cose
Reply to: