Scaling Icinga2 with many heterogeneous projects

on YouTube ↗47m

" summary

Managed hosting runs Icinga2 across two Berlin data centers monitoring ~3,000 Linux hosts and deploying 60,000 checks via Puppet, with Puppet runs optimized from ~15 to ~5 minutes. Full-state migration from Icinga1 preserves acknowledgments, downtimes and comments using the in-house tool icingadiff (soon on GitHub). Cluster uses two masters and two satellites per DC, NRPE for checks, events exported to Elasticsearch, and Icingaweb2 extended with an SLA marker and Zendesk ticket creation for incident workflow.

AI-written summary · about

" chapters (10)

1. 0:00Introduction and motivation
2. 6:26Requirements for new monitoring
3. 11:23Automating configuration with Puppet
4. 17:29Service templates and defaults
5. 20:21High availability setup
6. 26:26Full state migration challenges
7. 30:19Customizing Icinga Web
8. 35:21Lessons learned and tips
9. 37:10Is Icinga 2 perfect?
10. 38:32Q&A session

" related talks

• 46mWhy favour Icinga overBernd Erk · Rootconf 2015
• 29mHow Companies use IcingaBlerim Sheqa · Icinga Camp Berlin 2022
• 42mIcinga2 AutomatedJens Link · FrOSCon 2016
• 27mMonitoring at Scale with IcingaDave Kempe · Icinga Camp Berlin 2022
• 38mCurrent State of Icinga DBEric Lippmann · Icinga Camp Berlin 2022
• 29mAuto Discovery with IcingaLennart Betz · Open Source Monitoring Conference 2025
• 58mWhy favour Icinga over Nagios?Markus Frosch · FrOSCon 2015
• 45mIcinga for Windows – Age of PowerShellChristian Stein · Open Source Monitoring Conference 2023