This year’s Check_MK Conference took place in Munich on the 3rd and 4th of May, and with 220 participants, it’s shown a significant increase in interest since the last conference in October of 2015.

The program this year included a 5-part introduction of the new functionality developed in the past year, four guest presentations from customers and partners, dedicated discussions on the subjects of the new automatic agent updates and the Managed Services Edition, as well as a fascinating preview of upcoming developments. Our guest presenters discussed the subjects of business intelligence, security, end-to-end monitoring and monitoring in the enterprise environment.

At the beginning of the conference, Mathias Kettner introduced the entire Check_MK team and divulged a few key facts. The team has grown from 12 members at the time of the last conference in October 2015 to 20 today. 39 partners in 12 countries support the Check_MK team currently and account for 217 of the over 1000 subscriptions in use. Check_MK broke through the 1000 barrier around New Year‘s 2017 – an impressive demonstration of its popularity and market acceptance.

Following Mathias‘ presentation, the developer team provided some insights into the new functions added in the previous year. I’ll briefly describe some of these innovations here.



The number of data sources of an RRD file is automatically adjusted (by means of rrdtool tune) to changes in the number of sources of monitored performance data, without the need to change the storage settings of RRD files from SINGLE to MULTI. This saves the administrator from losing the data in his RRD files through manual deletion every time a change in the number of performance data sources occurs.

In addition to that improvement, administrators can now configure their own collections of multiple graphs using the function „custom graphs“. These can be saved, and even stored as a Python file in the backend. The graph data can also be extracted via an API in JSON format and, in future, will be available for saving without the need to deactivate monitoring. The processing of performance data using the rrdcache daemon can be paused and restarted using the „suspend“ and „resume“ commands.


Security in the „Agent Bakery“

There are innovations in the „Agent Bakery“ as well. One such is that user-defined files can be rolled out to the monitored systems. Another is that newly-created agents are signed, so that the updater within the Check_MK agent can recognize on the monitored system whether the intended update can be installed without fear of error. It can also recognize whether the update has been manipulated. Finally, the communication between the Check_MK server and the agents can be encoded.


Realtime Monitoring

As a result of improvements to the Check_MK MicroCore and the Check_MK agents, it is now possible to implement „realtime monitoring“. A separate process of the agent is forked to the monitored system and once there, transmits the desired data as a UDP packet to the MicroCore once a second.



The Web API has also received some new functionality. New calls can be used to start various functions such as Create, Delete, and Edit, thus allowing administrators to manage any object (such as devices, services, users, etc.) within Check_MK. Integrations can also be carried out in this way, for example the connection of a CMDB. And administrators will find the ability to call up a service directory particularly useful as it will open up tremendous new possibilities in the area of automation.


Event Console

The event console has been based on the Livestatus database since Version 1.4.0. This fact should appeal to all users in distributed environments. Messages from unknown devices, that is, devices which haven’t been registered in the monitoring system, can also be received and suitably handled. Additional new functionalities include the usage of regular expressions (RegExp) for filter mechanisms and „overflow protection“ which will prevent out-of-control floods of messages.


And there’s more! I’ll get to those other functionalities in the second part of this blog series next week. Subscribe to our blog as email and make sure you never miss another edition. (upper right)


Check_MK – what is that exactly?We explain the basic architecture and features of Check_MK in our Open Source Monitoring Guide.

We have in fact covered nine of the most popular open source monitoring solutions, comparing and contrasting each solution’s strengths and weaknesses. We hope you find it useful!


End of Part 1 – for Part 2 click here.