| Product scope | Prometheus is a metrics toolkit: the server scrapes and stores numeric time series, client libraries instrument code, and alert delivery and most dashboards come from separate components. | Zabbix is an integrated monitoring system that bundles data gathering, triggers, alerting, graphs, discovery, templates and an API in a single package. |
|---|
| Collection model | Prometheus pulls metrics over HTTP from instrumented applications and exporters. Short-lived jobs can push through an intermediary gateway. | Zabbix supports both polling and trapping. Agents on monitored hosts report to the server, and the server or proxies also run SNMP, IPMI, JMX and VMware checks. |
|---|
| Data model and queries | Each time series is identified by a metric name and key-value labels, and data is queried with PromQL for graphs and alert rules. | Zabbix is organised around hosts, items and triggers: an item is a metric collected from a host, and a trigger is an expression that defines a problem threshold. |
|---|
| Storage and retention | Prometheus writes to a local on-disk time series database with a default retention of 15 days. Local storage is not clustered or replicated, but remote storage integrations are available. | Zabbix stores all configuration and collected data in a database, with configurable history and a built-in housekeeping procedure. |
|---|
| Alerting and notifications | Alert rules run in the Prometheus server, and a separate Alertmanager handles silencing, inhibition, aggregation and notifications to email, on-call systems and chat platforms. | Zabbix triggers fire on thresholds in collected data, and actions send notifications with escalation schedules, recipients and media types, or run remote commands. |
|---|
| Discovery and templates | Scrape targets come from service discovery or static configuration, and the metrics available depend on the exporters and client library instrumentation you deploy. | Zabbix discovers network devices, file systems, interfaces and SNMP OIDs, supports agent autoregistration, and applies templates that can inherit other templates. |
|---|
| Dashboards and interface | Prometheus supports several modes of graphing and dashboarding, and its documentation names Grafana or other API consumers for visualising the collected data. | Zabbix graphs monitored items immediately and handles reports, statistics and configuration in its own web-based frontend. |
|---|
| Scaling and availability | Each Prometheus server is standalone and does not depend on network storage. Federation lets one server scrape selected time series from another. | Zabbix runs several servers as a high availability cluster with one active node, and proxies collect data on behalf of the server. |
|---|