Datadog Network Monitoring

The honest take

Datadog Network Monitoring isn’t really a product you buy — it’s a module you switch on once Datadog is already your everything. That framing matters, because it explains both why people love it and why they end up shopping for the exit. If your infra, APM, logs and Kubernetes already live in Datadog, turning on Network Device Monitoring (SNMP) and Network Performance Monitoring (flow) so you can pivot from “the app is slow” to “the link between these two services is dropping packets” in the same pane is genuinely excellent. Nobody else correlates network with the rest of the stack as smoothly.

Standalone, it’s a different story. Buying Datadog to do SNMP is paying a platform tax for a commodity. A dedicated NMS — Zabbix, LibreNMS, even PRTG — will give you deeper topology, config backup and device-centric workflows that NDM, for all its polish, still treats as second-class. Datadog’s network products are a complement to observability, not a replacement for a network management system.

Then there’s the bill, which is the recurring horror story across the entire platform and the single most common reason teams land on this site. The mechanics are worth understanding: Datadog meters many products separately (per-host infra, per-GB logs with separate indexing, custom metrics priced per unique tag combination, APM spans, NDM devices), and bills hosts on a high-water mark — your monthly charge is set by your peak host count, not your average. Autoscale hard during a traffic spike and you can pay all month for capacity you used for an hour. Add a chatty Kubernetes label and your custom-metrics line can quietly detonate. The list price isn’t the problem; the unpredictability is.

Our own status-page data (in the block above) is the other half of the honest picture: this is a heavily-used platform that posts real incidents. Read that as a busy, mature service rather than a verdict — but factor it in.

So: stay if Datadog is your home and the correlation is worth it, and you’ve modelled the bill against your real peak usage. If you’re here because of the invoice, the question isn’t “what’s the Datadog of X” — it’s whether per-host SaaS is the right pricing model for your fleet at all. The Datadog alternatives guide sorts the escapes by why you’re leaving.