This is the hub for the PingLabz Zabbix series. It's nine posts on Zabbix 7.0.32 LTS, built in my home lab against Cisco IOS XE 17.18.2 routers and switches and a couple of Ubuntu 24.04 servers. Every post shows the real screens and output, including the faults I hit along the way.
It starts with getting hosts in. The first post adds IOS XE devices over SNMPv3 with the stock Cisco IOS by SNMP template and explains how to read auth failures, timeouts and engine ID problems. The second adds a Linux server with Zabbix agent 2 and covers passive versus active checks and the two faults everyone hits: blocked ports 10050/10051 and a hostname mismatch. The third puts it all on a Grafana NOC board using the Zabbix plugin and a read-only API token.
The next three posts are about alerts people will actually read. Trigger dependencies, flap detection and hysteresis cut one real link shutdown from six problems to two. Email alerts get a second escalation step for problems nobody has acknowledged, tested with a real link failure. Syslog and SNMP traps catch config changes and link-downs faster than polling, and I timed the difference.
The last three are about running Zabbix at scale. Network discovery, agent autoregistration and a small Python script using the API onboard hosts without the host form. A checklist covers what to alert on and what to ignore when a new Cisco device joins. The final post deals with database growth, retention, a restore-tested backup and the path to 8.0.
The posts share one Zabbix server and one set of devices, so later parts build on earlier ones.

The series in reading order
- Cisco IOS XE in Zabbix over SNMPv3: Device Config, Template, and Auth Failure vs Timeout. Routers and switches over SNMPv3 with the stock template, and how to read auth, timeout and engine ID faults.
- Zabbix Agent 2 on Ubuntu: Passive vs Active Checks and the Three Faults That Break Them. Agent 2 on Ubuntu, passive versus active checks, and the port and hostname faults.
- Zabbix NOC Dashboard with the Grafana Plugin: Why a Shut Port Raised No Alarm. The Zabbix plugin for Grafana with a read-only token, and a kiosk NOC board tested with real failures.
- Zabbix Trigger Dependencies and Hysteresis: One WAN Failure from Six Problems to Two. Dependencies, flap detection and hysteresis that cut one outage from six problems to two.
- Zabbix Email Alerts with Escalation: The Media Severity Filter That Blocked Step 2. Media types, tag conditions and a second step for unacknowledged problems, tested with a real link failure.
- Zabbix Syslog and SNMP Traps from Cisco IOS XE: Trap Alert in 2.5 Seconds, Polling in 8.5. rsyslog with logrt[] for config changes, snmptrapd for link-downs, timed against polling.
- Zabbix Network Discovery, Agent Autoregistration, and Idempotent API Onboarding from a CSV. Network discovery, agent autoregistration and CSV onboarding through the API.
- Zabbix New Cisco Device Checklist: Which Template Triggers to Keep, Tune, or Ignore. What to alert on and what to tune when a new Cisco device joins.
- Zabbix Housekeeping and Backups: Measure Growth, Prove the Restore, Plan the 8.0 Upgrade. Database growth, retention, a restore-tested backup and the path to 8.0.
Where to start
Who this is for: anyone running Zabbix for network gear, or about to. Start with part 1 for a new install. If Zabbix already pages you too much, go straight to part 4.