Skip to content
Copy
View as Markdown Suggest changes
Add Docs MCP
Setup guide

Wallarm Daemon (wd) Metrics of the NGINX Node

The wd (Wallarm Daemon) service manages the NGINX Node's processes. This page describes the wd service's own Prometheus metrics.

What the wd service does

The wd service is the node's supervisor and is responsible for:

  • Process supervision — starts the node's services (wstore, wcli, API Firewall), monitors their state, and restarts them on failure.

  • Node registration and synchronization — registers the node with the Wallarm Cloud and continuously synchronizes its configuration and attack-detection resources (custom ruleset, proton.db, IP lists, IP feeds, and API specifications), reloading them as they change.

  • Metrics aggregation — collects the metrics of all managed services and serves them on a single aggregated endpoint (http://localhost:9445/metrics by default).

  • Health reporting — exposes readiness and liveness endpoints.

Limitations

wd metrics are not yet available for Sidecar.

Metrics endpoint

Starting from NGINX Node 7.1.0, the wd metrics are returned on the following aggregated endpoint, along with the other Node metrics:

http://127.0.0.1:9445/metrics

To change the port, see Configuring the ports.

Process supervision metrics

The wd service exposes its own wallarm_wd_* metrics that describe how it supervises the node's processes — as opposed to the functional metrics that each service emits about its own work:

  • wallarm_wd_up and wallarm_wd_start_timestamp — supervisor availability and start time.

  • wallarm_wd_managed_processes — number of managed processes.

  • wallarm_wd_process_state, wallarm_wd_process_restarts_total, wallarm_wd_process_uptime_seconds, and wallarm_wd_process_last_exit_code — per-process state, restarts, uptime, and last exit code (labeled by process name).

  • wallarm_wd_scrape_errors_total — errors encountered while scraping the managed processes.

  • metrics_push_* — counters for pushing metrics to the Wallarm Cloud (for example, metrics_push_total, metrics_push_errors_total, and metrics_push_bytes_pushed_total).