ITWorx Pulse brings Unraid host, container, storage, network and service telemetry into one operator-focused interface. It helps teams see what needs attention, investigate the underlying resource, and keep dashboards, alerts, incidents and wallboards in the same context.

Current release: v1.6.0. See the changelog for release history.

Pulse is operationally read-only: it observes and explains infrastructure, but does not start, stop, delete or repair workloads. Missing or stale telemetry appears as Unknown, never as a reassuring green status.

See Pulse in action

This capture uses isolated validation fixtures and does not show a production host.

Capabilities

Area	What operators can do
Overview and search	Move from an attention item to the exact host, container, service or storage resource.
Infrastructure detail	Inspect hosts, processes, containers, applications, arrays, disks, pools, shares and networks.
Dashboards	Compose versioned desktop, tablet, mobile and wallboard layouts.
Metrics	Explore bounded historical and live Prometheus queries without writing PromQL.
Alerts and incidents	Manage pending and recovery states, suppression, maintenance, timelines, notes and ownership.
Access and recovery	Use Authentik/OIDC roles, audit trails, verifiable backups and clean-room restore procedures.

The global operator bar keeps time context, entity search, favorites, saved views and recent resources available while investigating. The interface supports responsive layouts and accessible keyboard navigation.

Try it locally

Requirements: Go 1.26.6, Node.js 24+, pnpm 10.33+, Python 3, PowerShell 7 and Docker Compose.

Copy-Item .env.example .env
pwsh -NoProfile -File scripts/bootstrap.ps1
docker compose -f deploy/compose.yaml -f deploy/compose.dev.yaml up --build

Open http://localhost:18080. The development stack uses explicit mock authentication and isolated local data. Do not put production credentials or production telemetry in this environment. Stop the stack with docker compose -f deploy/compose.yaml -f deploy/compose.dev.yaml down --volumes when its local data can be discarded.

Architecture

flowchart LR
    Browser[Browser and wallboard] --> Web[Pulse web and API]
    Web --> DB[(PostgreSQL domain state)]
    Web --> Metrics[Prometheus metrics]
    Worker[Pulse worker] --> DB
    Agent[Read-only Unraid agent] --> Worker

The web/API tier has no Docker socket or host-root access. Only the agent receives explicitly configured read-only sources needed for observation. PostgreSQL stores configuration, inventory, dashboards, alerts and audit; the worker handles discovery, probes and notifications.

Read a board with its evidence

Start with the local mock environment to learn the board and incident views. In a real installation, configure the intended telemetry sources, then inspect timestamps and coverage alongside each status. Follow an incident into its supporting observations before deciding which infrastructure action to take.

Unknown or stale observations need investigation; they are not a healthy measurement. A running Pulse container does not prove that its remote telemetry sources are reachable. Pulse provides read-only operational evidence: perform infrastructure changes through the appropriate operator workflow.

Production deployment

Begin with the public deployment guide. A production instance needs HTTPS, Authentik/OIDC, private PostgreSQL, externally managed secrets, a defined Prometheus source, least-privilege Unraid API access where required, a backup directory, and a reviewed network and mount projection. The supplied configuration fails closed when mandatory production settings are absent. Validate backup, restore and rollback before relying on alerts.

Run the public source verification with:

pwsh -NoProfile -File scripts/public-verify.ps1

This checks Go tests and vet, frontend tests and build, contracts, wiring, secret markers, image digests and the public-source boundary. The optional Docker integration smoke is in scripts/integration-smoke.ps1.

README.md

ITWorx Pulse