my client uses an winapplication which is launching shell-scripts remotely on a HP-Unix Machine via Rexec.
The application-configuration is launching the script (which is in the home directory of connecting user) like:
So far so good, everything works.
Now they decided to get rid of the quite expensive HP-Unix server and asked me to do the same on a Linux-Distro. I set up CentOS and the REXEC-daemon, but....
1. When I launch the script against new Linux server with (same command as it is done towards old HP Unix):
it does not work (xinetd gives exitcode 127 : file not found)
2. When I launch the script with:
it works, but the problem is that i am not allowed to change the settings in the application. So I have to find a solution to launch the script like in point 1 above.
I tried already to put the script-folder to the PATH-variable, but it did not help.
Does somebody know what else I can do? My only little goal is to get this script launched using the commands at point 1.
PS: I know REXEC is not secure, but it's a requirement by company
In http://forums.unix.com/showthread.php?threadid=391
there is one statement called expect.
but when I issue command whereis expect, respond from o/s only EXPECT: (only one world).
I try to find it at /usr/bin, no expect statement there ?
is it default unix o/s command ?
I am using AIX on... (1 Reply)
Can someone tell me how I would enable Rexec on a UNIX machine? And is the procedure different on the different systems - Solaris, HP-UX -etc.
Thanks~!!
mike (1 Reply)
Rexec executes commands one at a time on a remote host. The rexec command provides an automatic login feature by checking for a $HOME/.netrc file. User and password are stored
in $HOME/.netrc.
I would like to log on to another host and execute a script/command but not using $HOME/.netrc file,... (4 Replies)
hi there,
i need some help, i am trying to run a script to launch a C program and a Java program but before running both I want to get a user input and then invoke both programs with input received. In the programs the inputs are not command line arguments.
This is the code,
after the java... (4 Replies)
Hi Team,
I am executing some ksh scripts which inturn calls java files in AIX Environment. We have installed java6_64 which is in .profile. But when we execute from rexec its taking path from some different place that does not have java in $PATH variable. Can you please help me find out which... (2 Replies)
Hi,
I am trying to enable rexec to automate certain tasks(it has to be rexec, not ssh or any other due to the system environment), so after switching to linux, I followed the certain instructions that were laid out in the web.
My operating system is fedora 17, so I first installed the... (1 Reply)
Hi all,
i have the following script
#!/bin/sh
for i in `ps -leaf --cols 1024 | grep LogUser | grep -v grep | awk '{print $4}'`;
do
echo $i
kill -15 $i;
done;
but it seems that the crontab its sciping this script,i configured corntab as following
*/30 * * * root... (2 Replies)
I'm using the Links2 console web browser in graphical mode (the "-g" argument), and launching a shell script
that invokes MPlayer from within it. MPlayer works fine. No problem there. The problem, is that I have no
control over the MPlayer process. I would like to be able to exit MPlayer whenever... (16 Replies)
hi all,
I created a sh script to import some tables from mysql to hive.
No problem launching it manually, but if I schedule via crontab it returns me an error in the following part:
#create an array containing all the tables for $dbname
query="SELECT table_name FROM information_schema.tables'... (10 Replies)
Discussion started by: mfran2002
10 Replies
LEARN ABOUT SUSE
rexec
REXEC(1) General Commands Manual REXEC(1)NAME
rexec -- remote execution client for an exec server
SYNOPSIS
rexec [ -abcdhns -l username -p password ] host command
DESCRIPTION
Rexec calls the rexec(3) routine to act as a client for the remote host's rexecd(8) server.
It asks that ``command'' be run on the host computer, using username/password authentication. See rexec(3) and rexecd(8) for details of the
protocol.
OPTIONS
Rexec accepts several options, but only three are likely to be very useful:
-l username
Set the log-in name on the remote host to username.
-p password
Provide the password for the remote account. The command line argument will be blanked after being parsed, to prevent it from being
seen with ps(1). However, it is still not very secure to type the password on the command line. In particular, be sure that the
shell's history file is protected.
-n Explicitly prompt for name and password, even if provided in the environment, in the $HOME/.netrc file, or in the environmental
variables REXEC_USER and REXEC_PASS.
Other options that might be useful with non-standard remote exec daemons, or to debug connections:
-a Do not set up an auxiliary channel for standard error from command; the remote standard error and output are then both returned on
the local standard output. By default, rexec asks that a separate channel be set up for diagnostic output from the remote command.
-b Use signal handling as in BSD rsh(1). Only the signals SIGINT, SIGQUIT, and SIGTERM are echoed to the remote process. They do not
remain raised locally, so rexec waits for the remote command to shutdown its side of the socket. Also, CNTRL-Z will only suspend
execution locally--the remote command may continue to run.
-c Do not close remote standard input when local standard input closes. Normally the standard input to the remote command is closed
when the local standard input is closed.
-d Turn on debugging information. In particular the command sent to the remote host will be echoed.
-h Print a usage message.
-s Do not echo signals received by the rexec onto the remote process. Normally, signals which can be trapped are passed on to the
remote process; then, when you type CNTRL-C, the remote process terminates as well.
USERNAME AND PASSWORD Rexec(1) searches for the username and password in the following order:
1. If -n is given on the command line, the user will always be prompted for both, even if they are also given on the command line.
2. The command line will be parsed
3. If the environmental variables REXEC_USER or REXEC_PASS are defined, they will define the username or password.
4. The $HOME/.netrc file will be searched. See ftp(1) for a description of this file's format.
5. Finally, the user will be prompted if either the username or password remains undefined.
SECURITY
Users of this command should be aware that rexec(3) transmits their password to the remote host clear text, not encrypted. If the network
is not secure to the remote host, the password can be comprimised.
SIGNALS
Without the -b option, all signals which can be handled are echoed to the remote process. Afterwards, however, they remain raised in the
local process. Typically, this means that rexec(1) will exit after receiving a fatal signal, even if the remote process has arranged to
handle or ignore it.
Differing operating systems use differing signal numbers; for example AIX and SunOS use 18 for SIGTSTP (^Z), while Linux uses 20. There-
fore, it may have a different effect remotely than locally. In particular, typing CNTL-Z may not suspend the execution of the remote
process.
EXAMPLE
rexec othermachine cat ">remote_file; date" <local_file
will send local_file to the othermachine as remote_file.
BUGS
Please send bug reports, system incompatibilities, and job offers to the author.
SEE ALSO rexec(3), rexecd(8), rsh(1)AUTHOR
Michael Sadd
mas22@cornell.edu
http://www.tc.cornell.edu/~sadd/
Thanks to Orange Gopher (2/10/97) and Johannes Plass (plass@dipmza.physik.uni-mainz.de, Oct. 17 1996) for useful suggestions.
February 14, 1997 REXEC(1)