Zabbix and Millions of Metrics: Best Practices for Large-Scale Monitoring
summary
Zabbix used to collect and manage millions of monitoring metrics; the talk examines challenges around automation, scalability, and integration with other systems when handling that scale. It discusses approaches for automating as much as possible so that large-scale monitoring can be set up once and then run with minimal ongoing intervention.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:25What is Zabbix and its architecture
- 4:58Zabbix Agent usage
- 7:04Metrics triggers and templates
- 11:24Low-level discovery
- 13:10Scaling and performance
- 16:02Database selection
- 21:18Data partitioning
- 23:02Liel Networks monitoring example
- 26:17Zabbix automation
- 29:30Magnit distributed monitoring
- 33:31Zabbix updates and compatibility
- 35:36Summary and recommendations
- 37:40Q&A
related talks
58mZabbix: Recipes for High-Performance Monitoring
34mZabbix: 100k NVPS on a Single Server
53mZabbix 3.2 — Monitoring on a Whole New Level
37mMonitoring as code: a GitOps approach with Zabbix
57mHow to Display 100,500 Metrics of a Distributed System and Not Go Insane
59mMonitoring A-Z
41mMonitoring Big Data Systems Done "The Simple Way"
42mMetrics at scale