Zabbix
Zabbix is an enterprise-class open source monitoring platform for hosts, networks, applications, and services. It polls agents and traps events, then surfaces alerts and dashboards from a web UI. We hold it for greenfield work: OpenTelemetry plus Grafana fit cloud-native, service-centric estates better than host-first monitoring.
Blurb
Zabbix is an enterprise-class open source distributed monitoring solution.
Summary
What it is: Server, agents or agentless checks, web frontend, alerting, and reporting over a central metadata database. Strong at SNMP, classic server metrics, and all-in-one ops teams that want one box.
When to use: Existing Zabbix estate with templates and on-call runbooks already paid for. Heavy SNMP network and datacenter monitoring where migration cost dominates.
When to skip: New SaaS or Kubernetes platforms where hosts are fungible and service-level signals matter more than per-host ping checks. Prefer Grafana with Prometheus, Loki, or cloud APIs, instrumented via OpenTelemetry.
Key features: Triggers and escalations, discovery, maps, SLA reporting, distributed proxies, and a large template library.
Details
| Topic | Notes |
|---|---|
| Deploy | Server plus DB (MySQL/Postgres/TimescaleDB), agents on hosts, optional proxies at the edge |
| Data model | Host-centric items, triggers, and graphs; service maps added over time but mindset stays host-first |
| Auth | Role-based UI access; LDAP and SSO in commercial tiers |
Practices: Treat Zabbix as legacy carry-forward, not the default observability stack. If you keep it, isolate read-only monitoring networks and cap alert noise with maintenance windows.
References