Page 1 of 1
Recommended monitoring for older RedHat releases?
Posted: Fri Sep 05, 2014 3:37 pm
by PhilG
Hello:
We are using Nagios XI 2014R1.4.
We have a few older Linux distros, RedHat 9 and RHEL 3, that will not be upgraded for quite some time due to our customers and timelines, etc.
Could the NRPE agent still be installed and used for monitoring on those Hosts?
Do you have any other recommendations besides the Bulk Host Wizard?
Thank you.
Re: Recommended monitoring for older RedHat releases?
Posted: Fri Sep 05, 2014 3:42 pm
by BanditBBS
You may have ot d/l and use and older version of nrpe, but it should work.
I use check_by_ssh here for all my linux checks and it works out well. I have a script I use that connects to the remote(customers) servers and determines the OS and copies the proper version of the plugins to the server.
Re: Recommended monitoring for older RedHat releases?
Posted: Fri Sep 05, 2014 4:09 pm
by PhilG
BanditBBS wrote:You may have ot d/l and use and older version of nrpe, but it should work.
I use check_by_ssh here for all my linux checks and it works out well. I have a script I use that connects to the remote(customers) servers and determines the OS and copies the proper version of the plugins to the server.
Would you care to share that script, por favor??
What version of NRPE would you recommend and is it available somewhere (link?)?
Thank you.
Re: Recommended monitoring for older RedHat releases?
Posted: Mon Sep 08, 2014 9:05 am
by BanditBBS
http://sourceforge.net/projects/nagios/files/nrpe-2.x/
I would definitely try the latest and see if it compiles and works. if not, then try an older version at that link.
If you aren't against using check_by_ssh instead of nrpe I will put my scripts here. Just keep in mind, the scripts assume you have ssh keys already in place for the nagios user and that /usr/local/nagios already exists with proper permissions for the nagios user(the OS team here makes sure of that with some script they wrote). I also have four folders in the folder with my script, they are:
1.) plugins - This folder has plugins that it doesn't matter the OS version or bit
2.) plugins32
3.) plugins64
4.) pluginsrh4
Notes that need mentioning:
1.) No making fun of my crappy coding

2.) If the urge is too strong, refer to #1
3.) You can easily just d/l and compile and install the nagios plugins. We do it this way so there is nothing that needs "installed" on the machines, just copied to them.
cdm.sh:
Code: Select all
#!/bin/sh
#set -e
#set -u
cat <<- EOF
#######################################
# #
# This is the installer script #
# for installing Nagios monitoring #
# plugins #
# #
#######################################
EOF
echo "Please enter the FQDN of the new Host"
read hostname
echo "You entered: $hostname"
# Connect to remote host
ssh nagios@$hostname 'bash -s' -- < /home/nagios/newservers/itcremote.sh
# copy from Nag04 to new host
RELEASE_VER=`/usr/local/nagios/libexec/check_by_ssh -H $hostname -C "cat /etc/redhat-release"`
if [[ "$RELEASE_VER" == *"release 4"* ]]; then
echo "release 4"
scp ~/newservers/pluginsrh4/* $hostname:/usr/local/nagios/libexec
scp ~/newservers/plugins/* $hostname:/usr/local/nagios/libexec
echo ""
echo "Folders created and plugins copied"
exit
fi
MACHINE_TYPE=`/usr/local/nagios/libexec/check_by_ssh -H $hostname -C "uname -m"`
if [ ${MACHINE_TYPE} == 'x86_64' ]; then
# 64-bit stuff here
echo "64 bit"
scp ~/newservers/plugins64/* $hostname:/usr/local/nagios/libexec
else
# 32-bit stuff here
echo "32 bit"
scp ~/newservers/plugins32/* $hostname:/usr/local/nagios/libexec
fi
scp ~/newservers/plugins/* $hostname:/usr/local/nagios/libexec
echo ""
echo "Folders created and plugins copied"
remote.sh: (also shows us the vnc listening ports, you can comment out that section)
Code: Select all
#!/bin/sh
if [ ! -d /usr/local/nagios/libexec ]; then
mkdir -p /usr/local/nagios/libexec;
fi;
chown -R nagios:nagios /usr/local/nagios
echo ""
echo "Here is VNC information"
echo ""
ps -ef|grep vnc
echo ""
Re: Recommended monitoring for older RedHat releases?
Posted: Mon Sep 08, 2014 11:57 am
by lmiltchev
Thanks, BanditBBS!
@PhilG
Let us know if BanditBBS's post helped.
Re: Recommended monitoring for older RedHat releases?
Posted: Wed Sep 10, 2014 6:23 pm
by rajasegar
PhilG wrote:Hello:
We are using Nagios XI 2014R1.4.
We have a few older Linux distros, RedHat 9 and RHEL 3, that will not be upgraded for quite some time due to our customers and timelines, etc.
Could the NRPE agent still be installed and used for monitoring on those Hosts?
Do you have any other recommendations besides the Bulk Host Wizard?
Thank you.
I just created a package for RHEL4 using nrpe 2.15 and nagios plugins 2.03
Just download and compile them in RHEL4
All the checks work just fine so far.
Re: Recommended monitoring for older RedHat releases?
Posted: Thu Sep 11, 2014 9:31 am
by slansing
How did the script work out for your Phil? We'll be closing this thread if all is good!
Re: Recommended monitoring for older RedHat releases?
Posted: Thu Sep 11, 2014 9:56 am
by PhilG
slansing wrote:How did the script work out for your Phil? We'll be closing this thread if all is good!
Bandit's information looks great, but the goal provided by a colleague here is to implement NRPE.
Additionally, I have another post asking about mass deployment/push of those clients to numerous Linux servers, and it appears to be something not easy (like SCCM for Windows).
Re: Recommended monitoring for older RedHat releases?
Posted: Thu Sep 11, 2014 10:36 am
by slansing
Well you can use Chef/Puppet, which are pretty much the excepted alternatives to SCCM for *nix. There is a boat load of documentation on doing what you are asking with them, I had just hoped we had our solution to this deployed by now.
Re: Recommended monitoring for older RedHat releases?
Posted: Thu Sep 11, 2014 10:50 am
by PhilG
slansing wrote:Well you can use Chef/Puppet, which are pretty much the excepted alternatives to SCCM for *nix. There is a boat load of documentation on doing what you are asking with them, I had just hoped we had our solution to this deployed by now.
That's okay. I understand R/D takes time, and I'm sure it's not an easy thing to accomplish. The effort your team puts in will be quite welcomed once the "features" have been worked out and the solution is provided for production to the masses!
Thanks, again! You may freeze this post.