Centreon Engine 1.4.0 – Call for contributor

This post is also available in: French

We are proud to announce the new Centreon Engine release (1.4.0 beta revision), our polling engine for Centreon.

So, we need all contributors to test all these new modifications. The task promises to be hard considering changes scale (more than 25 000 code lines have been changed).

We worked for several months on modifying the objects and indicators configuration system in Centreon Engine and finally, a new version is released. All this work was achieved to be able to avoid restarting all the engine each time the monitoring configuration is modified. Now we can make some modifications without any service downtime, this represents a revolution in Nagios environment.

Even if the old operating is well known, users looking for a 99,999% monitoring service availability will love this feature.

From now on, initialisation system takes into account the changes in configuration using the “reload” command. When Centreon receives a reload signal, it will analyse the configuration and apply the delta changes directly in memory. The engine won’t apply all the initialisation phases necessary to a restart, i.e.:

  • Stopping process :
    • send specific NEB flow data to stop the Centreon Engine instance (lots of SQL requests to reinitialize the whole perimeter)
    • retention data writing (waiting for .sav status writing)
  • Starting process :
    • configuration reading (data organisation)
    • retention data reading (reading of status.sav)
    • check scheduling queue regarding the .sav status file
    • send specific NEB flow data to start Centreon Engine instance (lots of SQL requests to reinitialize the whole perimeter)

All these actions can be time and resource consuming for poor results. It also creates an interruption in the monitoring. If configuration changes are limited, it is better to only make a “diff” and save time and resource by overcoming the instance initialization process and being able to use a more performant solution. Centreon Engine 1.4 applies all actions listed above on only modified areas and not on all Centreon Engine instance perimeter.

This new feature also corrects the interface “blink” effect when Centreon Enigne restart : all instance objects were disappearing few seconds and reappearing after the restart. This effect was feeling users uncomfortable, particularly when it concerns an important perimeter and when starting time can take up to several minutes.

We call you in order to constitute a good test team ! The version is being processed into our QA service but we are hoping from you many tests to validate the first release of 1.4.

If you are in, comment ! ;)

8 thoughts on “Centreon Engine 1.4.0 – Call for contributor

  1. Hi,

    that is a great feature !!!
    we have 4 system with nagios to monitor all our infrastructure and time to time we need apply change and the restart of nagios represent a downtime of 90 seconds (minimun) in each node !!!!

    We can put a centreon engine in a other node to try it.

    tell me, how we can participate ???

  2. I benefit from Centreon very much, and appreciate all your great efforts.

    The new innovation of new Centreon Engine is tremendous. I’m looking forward to contribute to the test. Please feel free to contact me.

Leave a Reply