netdata integration

Netdata Integration

You will learn how to export update-ipsets metrics to a local Netdata instance via OpenTelemetry.

Prerequisites

• Netdata installed and running with the otel-plugin enabled
• update-ipsets daemon running on the same host (or reachable over the network)

Netdata's otel-plugin listens for OTLP/gRPC on port 4317 by default.

Default systemd drop-in

The installed systemd unit already includes these defaults. If you customized the unit, add a drop-in at /etc/systemd/system/update-ipsets.service.d/otel.conf:

[Service]
Environment="UPDATE_IPSETS_OTEL=1"
Environment="UPDATE_IPSETS_OTEL_PROTOCOL=grpc"
Environment="OTEL_EXPORTER_OTLP_ENDPOINT=http://127.0.0.1:4317"
Environment="OTEL_METRIC_EXPORT_INTERVAL=10000"

Then reload and restart:

sudo systemctl daemon-reload
sudo systemctl restart update-ipsets

What this configuration does

• Enables OpenTelemetry metric export
• Uses gRPC protocol (required by Netdata's otel-plugin)
• Pushes metrics every 10 seconds, matching Netdata's default OTel chart interval
• Keeps logs local in bounded daemon-owned buffers; local traces stay disabled unless explicitly enabled with a trace buffer

Verifying the integration

After restarting, open Netdata and look for new charts under the update_ipsets application group. You should see counters for:

• Download operations (download.fetches, download.errors, download.fetch.bytes, download.fetch.duration_ms)
• Processing phases and queues (engine.runs, engine.run.duration_ms, scheduler.queue.*)
• Public/admin API activity (http.server.request.duration, api.recalculation.*)

Process CPU, memory, and file-descriptor usage is available through the admin status API under system and through Netdata's normal host/process monitoring charts.

If you don't see charts within 30 seconds, check:

1. Netdata is running: sudo netdatacli uptime
2. otel-plugin is enabled in netdata.conf
3. Port 4317 is open: ss -tlnp | grep 4317
4. Daemon logs show OTLP connection: journalctl --namespace=iplists -u update-ipsets -n 50

Tuning for Netdata

The 10-second metric interval aligns with Netdata's default OTel chart update rate. If you changed Netdata's update every for the otel-plugin, match that interval here:

# If Netdata's otel-plugin is configured for 5-second updates
Environment="OTEL_METRIC_EXPORT_INTERVAL=5000"

Shorter intervals produce finer charts but increase CPU and network overhead on both sides.

Remote Netdata

If Netdata runs on a different host, change the endpoint:

Environment="OTEL_EXPORTER_OTLP_ENDPOINT=http://netdata-host:4317"

Ensure the network path allows gRPC traffic on port 4317.