Public beta — temporary labs may change or be reset.

Checkmk setup guide

Validated

Tested with 2.3.0p50 Raw Edition

Checkmk discovers this device as an SNMP host. These steps mirror the M2E validation run.

Checkmk · SNMP discovery

Router (generic-router-v1)

SNMP v2c · 8 interfaces

Discovered
  • Uptime
  • 8 x Interface
  • CPU
  • Memory

Run high_cpu → CPU reads ~95% → your alert should fire.

Illustration of the expected discovery for a Router lab device, based on our 2.3.0p50 Raw Edition validation. Actual sensors depend on template, version and configuration.

Connection

Host
snmp-test.bonguides.com
Port
161
Version
SNMP v2c
Community
generated per device

The community is generated per device and shown only in your lab — treat it as a throwaway secret and do not reuse it. SNMP v2c is cleartext.

1. Create the device

  1. In the lab, create a Router (generic-router-v1) and copy its Host, Port, SNMP version and Community.
  2. Host is snmp-test.bonguides.com, port 161, SNMP v2c. The community is generated per device — treat it as a throwaway secret.
  3. Expect 8 interfaces (GigabitEthernet0/1–8), plus system uptime, CPU and memory over standard MIBs.

2. Configure Checkmk

Add the host

Create an SNMP host (no agent, SNMP v2 tag) with the address snmp-test.bonguides.com and the generated community, then run service discovery and activate changes.

Expected services

Discovery finds Uptime, 8 Interface services, CPU and Memory. Run cmk -v <host> to execute checks.

3. Run the fault scenarios

Trigger these from the device page in the lab, then watch Checkmk react. Each has a recovery action that restores the baseline.

  • interface_down — the interface 1 service reflects the down state.
  • high_cpu — the CPU service reads ~95%.
  • snmp_down — the SNMP check fails (CRIT/timeout), then normal monitoring resumes on recovery.

Caveats

  • Recovery resumes on the same host — you do not need to re-add the device after an snmp_down scenario.