Closed ghost closed 6 years ago
If you download the RPM packages and CPAN / PyPI dependencies then you can use any plugin without needing to run make
- all make
does is fetch the git submodules and install the system packages and CPAN and PyPI dependencies for Perl and Python respectively.
If building manually just install the packages for the plugin and make sure you have the lib
or pylib
git submodules on disk for Perl or Python respectively.
As this is quite a large project and there are a lot of dependencies you may want to consider using the pre-build docker images which contain all the dependencies and is the best option in your case for all checks that execute remotely over the network (which is most of them). You can download a docker image from a machine that does have internet access and push the image to your internal docker registry from there.
DockerHub images are available here as documented in the README:
This is so very frustrating. I tried manually downloading all dependencies, but more dependencies of those dependencies keep springing up, and so on. Furthermore, even some of the existing Nagios plugins that come installed by default when installing Nagios are not working, as I describe here: https://stackoverflow.com/questions/48003431/how-do-i-get-nagios-plugins-to-work
And to make matters worse, I could not find a decent tutorial on the web on how to install a Nagios plugin in a system with no internet access. (I was asked to install Nagios and Ganglia just 2 days ago, so I am a novice in this kind of stuff, this sort of thing is not my area of expertise.) Could you please help me out by guiding me on just how to get this damn 'check_hadoop_namenode.pl' to get working? Like, what are the bare minimum dependencies required to install it, what command I can type to download all the requisite dependencies, how to configure it in 'services.cfg', etc. It would be a great help! A very brief, high-level, step-by-step to-do list would be sufficient. Thanks.
The way I understand it, I need to install all dependency packages that are required to install 'check_hadoop_namenode.pl', and then define a service in /usr/local/nagios/etc/objects/services.cfg as follows:
define service{
use ganglia-service
service_description check_hadoop_namenode
check_command check_ganglia!check_hadoop_namenode!200!50!1
contact_groups admins
}
Or something along those lines. Am I right? Or do I need to do something more?
There should be no check_ganglia!
in front of check_hadoop_namenode and also you'd need a command definition. At the bottom of the main page are the links to the nagios documentation you need for service and command definitions.
The dependencies are basically like for every other Perl or Python program in the world, for which there are a million web pages describing, but if you have Docker that is the best option as you can download and run the self-contained docker image with a single command docker run harisekhon/nagios-plugins <program> <args>
.
I'm going to close this ticket as this isn't a problem specific to this project but general systems administration.
My system has no internet. I just want to use the service
check_hadoop_namenode.pl
. How can I install this plugin alone please? When I tried to follow your GitHub instructions, I cannotmake
as it weirdly says that several packages that are installed by default (like,which
,tar
,gcc
,whet
, etc) are not installed! A screenshot of a part of the error is attached below:I manually downloaded and installed the RPMs of each of the 30 odd packages it said was missing. But still, I am getting the same "missing package" error when
make
ing the clone of your GIT repo.Or, is there any way in which I can download and install all the dependency packages (and their dependencies, and so on) that are required when
make
ing, so thatmake
does not give error and runs successfully?