Features
Six topics: discovery and inventory, topology, alerting and integrations, the AI assistant, security and access control, and operations and deployment. Everything below ships in NETMapper Core under AGPL-3.0, and nothing in Core is time-limited or capped.
Discovery and inventory
NETMapper logs in to your switches over SSH and reads what they already know, so the inventory is what the network reports rather than what a spreadsheet remembers.
- Network inventory: MAC table, ARP table, hostname, vendor OUI, VLAN, port, and first/last-seen timestamps for every device on every switch.
- Scheduled scans: per-switch or scan-all-switches intervals via APScheduler; a manual scan is always available.
- DHCP integration: lease correlation via Kea, ISC dhcpd, MikroTik RouterOS, Windows DHCP (SSH and WinRM), pfSense, and OPNsense.
- Statistical anomaly detection: a daily batch scorer for device churn, IP churn, VLAN outliers, and unknown-device flags. Pure standard library, no ML dependencies.
- Eleven vendor scanners, all shipping in Core. The full table is further down this page.
Topology
The map is drawn from LLDP/CDP neighbor data collected on every scan, so it follows the cabling instead of a diagram somebody last edited two years ago.
- LLDP/CDP neighbor edges parsed from switch output and turned into a live topology graph.
- Interactive canvas with zoom, pan, and export.
- Follow MAC mode to trace one address across the fabric.
- Bookmarked topology views persisted server-side, so they follow the user rather than the browser.
Alerting and integrations
Changes in the inventory become alerts, and alerts go wherever your team already looks. Every delivery attempt is recorded.
- Alert rules: a flexible rule engine that fires on new device, device missing, vendor, VLAN, or custom field conditions.
- Channels: email (SMTP), Slack, Teams, HMAC-signed generic webhook, Telegram, Discord, Mattermost, PagerDuty, and OpsGenie.
- Per-rule throttling, per-channel quiet hours, and a delivery audit row for every attempt.
- REST API and Prometheus metrics for wiring NETMapper into the rest of your tooling.
AI assistant
A read-only assistant on top of the inventory NETMapper already collected. It answers questions about your data; it does not perform discovery and it cannot change anything.
- Chat-completion loop via any OpenAI-compatible provider (OpenAI, Anthropic, Google, Groq, OpenRouter, Azure/Copilot, Ollama, local).
- Tool use gated to five read-only lookups: list_devices, list_switches, list_neighbors, list_alerts, and stats_summary.
- AI keys are encrypted at rest, and conversations export as Markdown or JSON.
- Cmd+K command palette: navigate to any page, run a scan, toggle theme, log out. Typing ? or ai: routes the query straight into the AI chat.
Security and access control
NETMapper holds switch credentials, so the access model and the encryption are part of the product rather than an add-on.
- Role-based access control: a 24-permission catalog across 9 categories with three built-in roles (admin, operator, viewer). Custom roles are a Pro feature.
- Two-factor authentication with TOTP.
- HttpOnly cookie plus CSRF double-submit on every mutation; the access token never lives in localStorage.
- Encryption at rest: switch credentials, notification-channel configs, and AI keys are sealed with Fernet, backed by your own encryption key.
- bcrypt password hashing with a Have I Been Pwned k-anonymity check, and rate-limited login endpoints.
- Audit log with configurable retention: every privileged and mutating action writes a persistent record. Retention defaults to 90 days, with a daily prune job and on-demand pruning from the UI.
- Policy acknowledgement: publish your own acceptable use policy, terms, or privacy notice and track per-version acceptance per user.
Operations and deployment
One Docker Compose command is the short path. Kubernetes, bare metal and the day-to-day surfaces are all covered too.
- Docker Compose install with digest-pinned images, plus built-in HTTPS via nginx with automatic Let's Encrypt certificates.
- Helm chart (beta) for Kubernetes, and a bare-metal install script with a systemd unit.
- Prometheus metrics endpoint, plus health and readiness endpoints.
- Installable PWA shell with a service worker and an offline fallback page.
- Server-side user preferences: topology bookmarks, device tags, and dashboard widget order persist per user.
- Drag-and-drop dashboard widget reordering with a reset button, and a print stylesheet for management reviews.
Core vs Pro
Core, AGPL-3.0
Every topic on this page, for homelabs and small teams. Uncapped.
Pro adds
Syslog and EDR / XDR / SIEM channels, compliance evidence reports, custom roles, active-active replicas, and a Vault secrets provider, for businesses that need them.
Supported vendors and scanners
Eleven vendor scanners, all shipping in Core. Each registers itself with the scanner factory, so adding a vendor is a self-contained change.
| Vendor | Tested against | Notes |
|---|---|---|
| HPE Comware | A5500, A5120, 5900, 5130 | H3C silicon |
| HPE ProCurve | 2510, 2610, 2910, 3400, 5300 | Legacy ProCurve/Aruba |
| Cisco IOS / IOS-XE | Catalyst 2960, 3750, 3850, 9300 | |
| Cisco NX-OS | Nexus 3k/5k/7k/9k | Data center |
| Aruba AOS-CX | CX 6xxx/8xxx/10xxx | Newer Aruba |
| Aruba AOS-Switch | 2530, 2920, 2930, 3810, 5400 | Legacy HP/Aruba |
| Juniper Junos | EX series, QFX series | |
| Dell OS10 | S3000, S4100, S5200, Z9100 | |
| Extreme EXOS | X440, X450, X460, X670, X870 | |
| Nokia SR Linux | SR Linux via containerlab | Verified in lab |
| SONiC | SONiC VS via containerlab | Verified in lab |
FRR/Linux acts as an LLDP peer in the lab topology but does not have a dedicated scanner yet. Need another vendor? Open an issue. Adding scanners is one of the highest-value contributions right now.
REST API
The same API the dashboard uses. Interactive docs are served by every install at/docs (Swagger UI) and /redoc.
- Swagger UI and ReDoc documentation
- API tokens for automation, and JWT cookie auth for the browser
- Mutating endpoints gated by granular permission checks and written to the audit log
- Health and readiness endpoints, plus Prometheus metrics
A versioned, rate-limited public API is on the roadmap for v1.0.0.
/devicesList discovered devices
/devices/{id}Device detail
/neighborsLLDP/CDP neighbors
/switchesAdd a switch (admin)
/scanScan all active switches (async, 202)
/alertsList alerts
/ai/queryAI chat-completion query
/healthLiveness check
Ready to get started?
Read the docs, or clone the repository and run docker compose.