Raspberry Pi Monitoring for IBM i

Raspberry Pi single board computer

We mentioned previously about a project we undertook to add IBM i monitoring to Nagios that resulted in the use of a Raspberry Pi and the NEMS (Nagios Enterprise Monitoring Solution) distribution that is maintained on line by a fellow Canadian who resides close by. The following link takes you to that post NEMS and IBM i Monitoring. If you are interested in the NEMS project you can find it here. The solution was built entirely with Open Source including the IBM i plugin for Nagios which was developed by IBM which can be found at the following Link Github Nagios for IBM i. (We have recently published a post on how to fix up the wizard zip files for Nagios XI as they would not upload and install using the Nagios XI wizard installer. This post can be found Here.)

Most Enterprise level customers we have spoken with want to run Nagios Monitoring on a purpose built server or a VM due to the size of their infrastructure (some wont even consider having a Raspberry Pi in their network), but for the smaller customer that wants to add some level of monitoring without the expense and complexity of running additional Linux systems or VM’s the Raspberry Pi is a perfect solution.

We moved away from NEMS solution as the support was not responsive or effective, the IBM i plugin was also being supported out of China at the time (those developers have now gone) which resulted in long delays getting issues via Github resolved. However the initial tests of the IBM i plugin started to raise some concerns such as performance and the overhead both on the Linux side and the target IBM i partition for each check that ran. We looked through the code for the plugin in the hopes that we could improve the performance, but being written in Java and the design of the process made that impossible for us improve on. After a few attempts to get changes added or fixes implemented from the NEMs team and the IBM Github issues page failed to get resolved in a timely fashion we decided the NEMS solution was not for us.

Nagios XI is something we needed to stick with due to the number of customers who run it and our products, but it became too expensive for customers that really only needed to monitor a few IBM i systems and very little else. To fill this gap we built our own Nagios Core Monitoring Stack using the Raspberry Pi as the hardware platform, eventually building the exact same stack for Debian,Ubuntu and CentOS as a VM option which we maintain and ship to our customers. Some of the elements that made up the NEMS solution such as Nagios TV and Terminal impressed us with their support and simplicity so they became part of our own stack and a number of others such as Grafana for building graphs and our own PHP interface to simplify the configuration and management of the monitors.

Everything that can be run on Nagios XI can be run on the RPi stack that we provide, even the new Power HMC checks. Our internal systems are all monitored via the RPi implementation we maintain, we have yet to see an performance related issues running Nagios Core on the RPi or have any problems with the stability of the hardware and software. We monitor over 18 LPARs internally plus other infrastructure which results in 10,000’s of checks per hour, the RPi does all of this while taking minimal power.

If you are interested in the Shield RPi solution or would like to have a demo of our Nagios Plugins at work let us know, we will be more than happy to schedule a demo at your convenience.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.