We love monitoring here at Spearhead. It is the cornerstone of our service delivery methodology and a highly ingrained process that we could not live without but moreover would not be able to provide the quality and responsiveness of our services.

Before we engage in a project we first measure. By measuring we gather as much input as possible from the network, systems, and applications to set a baseline. From this baseline we can begin to design a system based on solid data. This is our preferred methodology, using a scientific method that helps us and our customers get things right the first time. By monitoring we do not have to base our recommendation on gut feelings, past experience extrapolated with new circumstances. We use science to get it right the first time.

Our preferred monitoring system is hands down Check_MK so it is of little wonder that we were present at the Check_MK Conference #3 .

Here is a short breakdown of what was presented on the first day of conference.

We got to meet the expanding Check_MK team and speak to the developers that we work with so closely over digital communication channels. Seeing and talking to them live was a pleasure.

Check_mk Team

Newest developments in Check_MK

Mathias kicked off this presentation by introducing some of the most interesting developments in Check_MK. In no particular order I jotted down the following:

  • Combined graphs

This new feature allows us to create graphs from any view showing a selection of hosts or services. These combined graphs will have all the metrics from the hosts services in a combined fashion so that you can see (for example) all HTTP- metrics from a web farm in a single combined graph. You can also choose a presentation format such as stacked, sum, min, max, average and lines.

Combined graphs

  • Graph designer

There is also a new graph designer allowing for easy creation of custom performance graphs from arbitrary metrics from hosts and services.

Performing complex calculations on your monitored data.

  • API for graph data

This one is handy. The Web API has the ability to fetch graph data. This allows us access to graph data via HTTP/JSON> The data is preprocessed, not just the bare RRD data and this works in distributed setups as well.

  • Session timeouts

Lots of werks relate to the UI and this particular feature allows for setting of global or per-user session timeouts. This is handy in environments where there may be strict policy enforcement for session activity.

  • Login limits

User logins can now be limited to only single session at a time. This basically means a user cannot log-in with the same username from two browsers.

  • Stale services

These can now be viewed directly in the tactical overview which makes it really easy to quickly get a complete overview of the entire IT infrastructure.

There were well over 100 werks, I jotted down those of most interest to me. Check out the official slides or give us a call and we’d love to dig into the details of the new features you can look forward to in Check_MK.

Next up was Fabian Binder (ComNET GmbH). He presented Using Check_MK BI for modelling IT services.

Check_MK’s BI is a real treasure. We use it to model complex IT services that help us gauge the real status of our services. It also allows us to run simulations to gauge impact if something were to happen to one of our services.

Fabian dug deep into application and end-user monitoring using open source components such as WebInject. If you’re interested in doing complex IT service modelling and end-user monitoring give us a shout and we can show you how we integrate common end-user browsers into check_mk for end-to-end monitoring.

Next up we got some details about the The new automatic agent updates

This feature is very interesting as it not only allows for the agents to automatically update themselves but they can also update plugins and more recently we can include custom files. This allows for the check_mk system to do considerably more such as deploying classical nagios plugins for use with MRPE or pushing any other file that may be necessary for monitoring. These new features work on both Linux and Windows.

Next we looked at part 2 of Newest developments in Check_MK

Here the Check_MK team covered real-time checks that can handle one second resolution. Integration with Graphite (and compatible DB’s such as InfluxDB) has been updated and all Check_MK metrics can be sent to a graphite server which in turn can be viewed with tools such as Grafana. There were a host of new features such as disabling the sending of metrics per service, sending of the warn/crit/min/max values along with the performance data and many more including Livestatus improvements that handle queries of up to 500MB.

Next up was Ralf Spenneberg who presented on Check_MK and Security

A very interested subject since recently Check_MK has been gaining more features that allows us to do more, such as pushing custom files to our monitored systems. He presented to us how to use Check_MK in a secure way by understanding the principles and concepts.

Finally we closed the day with Newest developments in Check_MK - Part 3

The Check_MK Team presented some details about the release cycle, naming conventions and how developing new features is handled. There were several werks presented here as well which I will leave up to you to discover by viewing the official slides and presentation or giving us a shout to provide a demo.

All of this was just the first day! We had a great second day and we are happy to let you know that we are one of the most active Check_MK partners in our region and soon will be expanding the new markets.

New markets.