SNMP monitoring

What it's for

SNMP monitoring lets you supervise network devices and servers by querying their SNMP counters: CPU load, memory, storage, temperature, interface throughput, wireless signal, sessions… SecuMon provides ready-to-use templates per device type, which group the right OIDs with human-readable names and default thresholds.

An "SNMP device" in SecuMon is a derived view: it corresponds to a check of type snmp attached to a host. You therefore create an SNMP device by creating an snmp check, then you view its metrics on the host page or through the API.

Available device templates

Specify a template in the template field of the SNMP check configuration. The template automatically determines the OIDs queried.

A few notable metrics and their default thresholds:

• linux: cpuIdle (warning at 20%, critical at 5%), hrStorageUsed (warning at 80%, critical at 95%), hrProcessorLoad (80% / 95%).
• cisco: cpmCPUTotal1min / cpmCPUTotal5min (80% / 95%), ciscoMemoryPoolUsed (80% / 95%), ciscoTemperature (60 °C / 75 °C).
• fortinet: fgSysCpuUsage and fgSysMemUsage (80% / 95%), fgSysSesCount (80,000 / 100,000 sessions).
• ubiquiti: ubntWlStatSignal (warning at −75 dBm, critical at −85 dBm).

If no template is specified and no explicit OID is provided, SecuMon queries a minimal set of standard OIDs (sysName, sysUpTime, sysDescr).

Create SNMP monitoring

From the portal

1. Create (or reuse) a host for the device (web-prod-01, 10.0.0.12).
2. Add a check of type snmp on this host.
3. Fill in the configuration: address, version, community and template.
4. Confirm. The metrics will then appear under SNMP Metrics on the host page, as well as in the SNMP section.

Creation reserved for the owner and admin roles.

Through the API

Create an snmp check (see also Checks):

curl -s -X POST https://api.secumon.secuaas.ovh/api/v2/checks \
  -H "Authorization: Bearer $SECUMON_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "host_id": "8f3b1c2a-2d44-4e1a-9b77-0a1b2c3d4e5f",
    "name": "snmp web-prod-01",
    "check_type": "snmp",
    "config": {
      "host": "10.0.0.12",
      "version": "2c",
      "community": "public",
      "template": "linux"
    },
    "interval_sec": 120,
    "timeout_sec": 10
  }'

SNMPv3

For a device using SNMPv3, specify the security level and the credentials:

{
  "host": "10.0.0.12",
  "version": "3",
  "security_level": "authPriv",
  "username": "monitor",
  "auth_protocol": "SHA",
  "auth_passphrase": "...",
  "priv_protocol": "AES",
  "priv_passphrase": "..."
}

Read the metrics

List SNMP devices

GET /api/v2/snmp-devices returns the inventory of your SNMP devices (derived from the snmp checks and their hosts).

curl -s https://api.secumon.secuaas.ovh/api/v2/snmp-devices \
  -H "Authorization: Bearer $SECUMON_TOKEN"

[
  {
    "id": "1a2b3c4d-5e6f-4a7b-8c9d-0e1f2a3b4c5d",
    "host_id": "8f3b1c2a-2d44-4e1a-9b77-0a1b2c3d4e5f",
    "name": "snmp web-prod-01",
    "hostname": "10.0.0.12",
    "metrics": {}
  }
]

The response is always a JSON array.

Time series for a host

GET /api/v2/hosts/:id/snmp-metrics?from=&to= returns the series of the host's SNMP checks, enriched with the human-readable name of each check. By default, the window covers the last 6 hours.

curl -s "https://api.secumon.secuaas.ovh/api/v2/hosts/8f3b1c2a-2d44-4e1a-9b77-0a1b2c3d4e5f/snmp-metrics?from=2026-06-23T06:00:00Z&to=2026-06-23T12:00:00Z" \
  -H "Authorization: Bearer $SECUMON_TOKEN"

{
  "checks": [
    {
      "check_id": "1a2b3c4d-5e6f-4a7b-8c9d-0e1f2a3b4c5d",
      "check_name": "snmp web-prod-01",
      "points": [ /* time points per metric */ ]
    }
  ]
}

OID names (e.g. hrProcessorLoad, cpmCPUTotal1min) are resolved into human-readable labels in the output, which spares you from interpreting the numeric identifiers.

Use cases

• Linux server: the linux template to track CPU, memory, swap and storage, with an automatic alert beyond the default thresholds.
• Router / firewall: the cisco, mikrotik or fortinet templates for temperature, load and interface state.
• Wireless link: the ubiquiti template to monitor signal degradation (ubntWlStatSignal).

Tips

• Start with a standard template rather than listing the OIDs by hand: it is faster and the thresholds are already tuned.
• For a specific need, provide an explicit oids array in the config: it takes priority over the template.
• A 120 s interval suits most devices; go lower only for metrics that change quickly.
• If you are testing SNMP connectivity, start with a simple numeric OID to validate the community/credentials before adding a full template.