I’ve installed a plug-in from Source Forge ( check_remote_process ) which is basically a Perl script which communicates to remote hosts via ssh. Is there a better way to check random services on remote hosts? Your feed back would be much appreciated.
Most people prefer nsca using distributed monitoring.
There is also nrpe, which is a bit more resource hungry. Check out the Nagios doco.
you could use an SNMP check… There are ready to go process checks in the standard snmpd.conf file. 
Luca
I use the check_by_ssh command to check remote servers and nsca to check remote hosts on foreign networks. Also I use the nc_net client for windows servers. All work for the specific purposes they were intended.
My one beef may be that the check_by _ssh command takes some time because the server has to login to the remote server and then get the request. NSCA is much faster in returning the checks, but then you have to configure nagios to run on the server, which is why I use check_by_ssh to check 5-8 services on 20-30 servers. I guess you could use cron jobs to run the commands and send them back to the nagios server, but that seems like alot of work, and check_by_ssh is pretty easy to implement.
here is a script I use to configure some services on a remote host
#copy check_* to remote server /root
useradd nagios
su nagios
done as nagios, the Key is the public key from nagios@nagios echoed into ~/,ssh/authorized_keys
mkdir /home/nagios/.ssh
echo ssh-dss > /home/nagios/.ssh/authorized_keys
chmod go-w ~/
chmod 700 ~/.ssh
chmod 600 ~/.ssh/authorized_keys
exit
#This is done as root
mkdir /usr/local/nagios
mkdir /usr/local/nagios/libexec
cp /root/check_procs_redhat /usr/local/nagios/libexec
chown nagios /usr/local/nagios/
chown nagios /usr/local/nagios/libexec/
chown nagios /usr/local/nagios/libexec/*
chmod 755 /usr/local/nagios/libexec/check_*
I guess the real question is what systems/services are you trying to monitor? There are many tools and if used properly they can be very powerful.
Good luck
Edited Fri Feb 03 2006, 08:23AM ]