SNMP Monitoring

SNMP (Simple Network Management Protocol) is the standard protocol for monitoring network devices — switches, routers, firewalls, UPS units, and hypervisors like Nutanix. Last9 collects SNMP metrics via the OpenTelemetry Collector or the Prometheus SNMP Exporter.

How it works

SNMP Device                  Collector                    Last9
(switch, router,  ──GET/WALK──▶  OTel Collector       ──OTLP──▶  Levitate
 Nutanix, UPS)                   or Prometheus                    (metrics)
                                 SNMP Exporter

The collector polls SNMP-capable devices at a configurable interval, converts the OID values to metrics, and forwards them to Last9.

Prerequisites

OpenTelemetry Collector Contrib v0.90.0 or later (for snmpreceiver)
Or Prometheus SNMP Exporter if you prefer the Prometheus path
Network access from the collector host to the SNMP device (UDP port 161)
SNMP community string (v1/v2c) or credentials (v3)
Last9 OTLP endpoint and authentication credentials

Approach 1 — OTel snmpreceiver (recommended)

Use this when you want a single OTel Collector to handle SNMP alongside other receivers (hostmetrics, sqlserver, etc.).

Install OTel Collector Contrib

Download from the releases page and install as a service. If you already have a collector running for host metrics, add the SNMP receiver to the same config.

Configuration

receivers:
  snmp:
    collection_interval: 60s
    endpoint: "udp://192.168.1.1:161" # Replace with device IP
    version: v2c
    community: "public" # Replace with your community string
    # For SNMPv3:
    # version: v3
    # user: "monitoruser"
    # security_level: auth_priv
    # auth_type: MD5
    # auth_password: "${env:SNMP_AUTH_PASSWORD}"
    # privacy_type: DES
    # privacy_password: "${env:SNMP_PRIV_PASSWORD}"
    resource_attributes:
      device.name:
        scalar_oid: "1.3.6.1.2.1.1.5.0" # sysName — scalar, not a table OID
    attributes:
      interface_name:
        oid: "1.3.6.1.2.1.2.2.1.2" # ifDescr — looked up by the same table index
    metrics:
      # Interface traffic — cumulative counters
      network.interface.in.octets:
        unit: "By"
        sum:
          value_type: int
          aggregation: cumulative
          monotonic: true
        column_oids:
          - oid: "1.3.6.1.2.1.2.2.1.10" # ifInOctets
            attributes:
              - name: interface_name
      network.interface.out.octets:
        unit: "By"
        sum:
          value_type: int
          aggregation: cumulative
          monotonic: true
        column_oids:
          - oid: "1.3.6.1.2.1.2.2.1.16" # ifOutOctets
            attributes:
              - name: interface_name
      # System uptime — gauge (resets on reboot)
      system.uptime:
        unit: "cs"
        gauge:
          value_type: int
        scalar_oids:
          - oid: "1.3.6.1.2.1.1.3.0" # sysUpTime
      # CPU (Cisco IOS example)
      system.cpu.utilization:
        unit: "%"
        gauge:
          value_type: double
        scalar_oids:
          - oid: "1.3.6.1.4.1.9.2.1.57.0"

processors:
  resourcedetection:
    detectors: [env, system]

exporters:
  otlphttp:
    endpoint: "https://<your-last9-otlp-endpoint>"
    headers:
      Authorization: "Basic <your-base64-credentials>"
    compression: gzip

service:
  pipelines:
    metrics:
      receivers: [snmp]
      processors: [resourcedetection]
      exporters: [otlphttp]

Monitor multiple devices

Add one snmp receiver per device, each with a unique name:

receivers:
  snmp/core-switch:
    endpoint: "udp://10.0.0.1:161"
    community: "public"
    # ... metrics config
  snmp/firewall:
    endpoint: "udp://10.0.0.2:161"
    community: "public"
    # ... metrics config

service:
  pipelines:
    metrics:
      receivers: [snmp/core-switch, snmp/firewall]
      processors: [resourcedetection]
      exporters: [otlphttp]

Approach 2 — Prometheus SNMP Exporter

Use this when you already have a Prometheus SNMP Exporter running, or prefer managing MIB files with the Prometheus toolchain.

Run the Prometheus SNMP Exporter

docker run -d \
  --name snmp-exporter \
  -p 9116:9116 \
  -v ./snmp.yml:/etc/snmp_exporter/snmp.yml \
  prom/snmp-exporter:latest

Generate snmp.yml from MIB files using the snmp_exporter generator.

Scrape with OTel Collector

receivers:
  prometheus:
    config:
      scrape_configs:
        - job_name: snmp
          static_configs:
            - targets:
                - 192.168.1.1 # Device IP — passed as 'target' param to exporter
          metrics_path: /snmp
          params:
            module: [if_mib]
          relabel_configs:
            - source_labels: [__address__]
              target_label: __param_target
            - source_labels: [__param_target]
              target_label: instance
            - replacement: localhost:9116
              target_label: __address__

exporters:
  otlphttp:
    endpoint: "https://<your-last9-otlp-endpoint>"
    headers:
      Authorization: "Basic <your-base64-credentials>"
    compression: gzip

service:
  pipelines:
    metrics:
      receivers: [prometheus]
      exporters: [otlphttp]

Nutanix cluster monitoring

Nutanix Prism exposes SNMP OIDs for cluster health. Use the snmpreceiver with Nutanix-specific OIDs, or use the nutanix-exporter which queries the Prism REST API and exposes Prometheus metrics.

# Nutanix via snmpreceiver
receivers:
  snmp/nutanix:
    endpoint: "udp://<prism-ip>:161"
    version: v2c
    community: "public"
    metrics:
      nutanix.cluster.cpu.usage:
        unit: "%"
        gauge:
          value_type: double
        scalar_oids:
          - oid: "1.3.6.1.4.1.41263.1.1" # clusterCpuUsagePpm — divide by 10000 for %
      nutanix.cluster.memory.usage:
        unit: "%"
        gauge:
          value_type: double
        scalar_oids:
          - oid: "1.3.6.1.4.1.41263.1.2"
      nutanix.cluster.storage.usage:
        unit: "By"
        gauge:
          value_type: int
        scalar_oids:
          - oid: "1.3.6.1.4.1.41263.2.1"

Common OIDs reference

Metric	OID	Notes
System name	`1.3.6.1.2.1.1.5.0`	sysName
System uptime	`1.3.6.1.2.1.1.3.0`	sysUpTime (centiseconds)
Interface in bytes	`1.3.6.1.2.1.2.2.1.10`	ifInOctets (table)
Interface out bytes	`1.3.6.1.2.1.2.2.1.16`	ifOutOctets (table)
Interface errors in	`1.3.6.1.2.1.2.2.1.14`	ifInErrors (table)
Interface errors out	`1.3.6.1.2.1.2.2.1.20`	ifOutErrors (table)
Interface status	`1.3.6.1.2.1.2.2.1.8`	ifOperStatus (table)

For vendor-specific OIDs (Cisco, Juniper, HP, Nutanix), refer to the device’s MIB files.

Approach comparison

	OTel snmpreceiver	Prometheus SNMP Exporter
Setup complexity	Low — single config file	Higher — MIB generator step
MIB support	Manual OID config	Full MIB file support
Best for	Known OIDs, quick setup	Large MIB libraries, Prometheus users
Existing setup	Works standalone	Needs Prometheus or OTel scraper

Verify metrics in Last9

After starting the collector, open Last9 → Metrics and search for your metric name (e.g., network.interface.in.octets). Filter by device.name to see per-device data.

Create a dashboard panel with:

rate(network_interface_in_octets{device_name="core-switch"}[5m]) * 8

This gives interface throughput in bits per second.

Troubleshooting

Symptom	Solution
No metrics in Last9	Check UDP port 161 is reachable from collector host: `snmpwalk -v2c -c public <device-ip> 1.3.6.1.2.1.1.5.0`
Authentication error	Verify community string or SNMPv3 credentials; check device SNMP ACL allows collector IP
OID returns no data	Confirm OID exists on device: `snmpget -v2c -c public <device-ip> <oid>`
Metrics have wrong values	Some OIDs are counters — use `rate()` in PromQL instead of raw values
High collector memory	Reduce `collection_interval` frequency or limit number of OIDs polled per device