The apps, one page each
Every app stands alone - its repository, README, and docker compose are self-contained. These are the short versions, from the suite's point of view. The ports are the stock ones; your tiles on the Launch page reflect any overrides.
The diagram editor - where the network gets drawn. A large stencil
library, zones, themes, and importers for Visio (.vsdx),
draw.io, Gliffy, and inventory CSVs. Saves to a plain-JSON
.xcanvas file that the rest of the suite treats as the source
of truth.
First five minutes
- Drag stencils from the sidebar, connect them, group with zones.
- Have an existing diagram? File → Import Diagram. Have a device list? File → Import Inventory.
- Give monitored devices an IP-Address in Device Details - that is the wall's whole contract.
- Pick a theme (29 of them); exports and the wall inherit it.
Your diagram as a live wall. A poller pings every device with an
IP-Address and the kiosk page recolors the diagram by reachability - green,
amber, red - and overlays live SNMP readings where the diagram carries
{code} tokens. Built for a TV that never gets a keyboard: no
login, URL-tunable, burn-in-safe theme rotation.
First five minutes
- Upload a board from the Launch page; the kiosk picks it up on its next refresh.
- Open the kiosk tile - that's it, it runs itself.
- Want different colors, refresh rates, or theme rotation? It's all URL parameters - the README has the table.
The numbers. Polls your devices over SNMP (v2c or v3) for interface
traffic, CPU, memory, temperature, fans, power, UPS state - with vendor
smarts for the gear it recognizes. Keeps history, draws graphs, and exports
the live readings (snmp-status.json) that the wall and
AlertCanvas consume.
First five minutes
- Devices → Add (or Bulk add): address + credentials; the probe shows what the device offers before you commit.
- Check Export on values you want on the wall - each gets a
{code}token, click to copy. - Settings → Inventory export turns the monitored fleet into a CrossCanvas-importable CSV - the fastest route to a first diagram.
The flight recorder. Listens for syslog messages and SNMP traps from your devices, stores them with sensible retention caps, and gives you a fast filterable view - severity, host, text - for the day something turns red and you need to know what it said beforehand.
First five minutes
- Point a device's syslog at this host (514/udp) - most network gear is one config line.
- Send traps to 162/udp; they land in the same stream, decoded.
- Watch the live tail, then try a filter - it's all server-side, so it stays fast at hundreds of thousands of rows.
Thresholds to notifications. Watches the same
snmp-status.json feed as the wall and raises alarms when a
value crosses a line you set - device down, CPU pegged, temperature
climbing, UPS on battery. Delivers by email (SMTP), ntfy push, or syslog,
with warn/crit levels and per-device overrides.
First five minutes
- Set global thresholds in Settings; override per device where a hot closet is just a hot closet.
- Add a channel (email, ntfy, or syslog) and use the built-in test to prove delivery end to end - including the 2 AM-style alarm mail.
- Device down alerts work out of the box for every exported device.
The front door - the app you're reading this in. One login for the
suite (with optional single sign-on into the Node siblings via a shared
SUITE_SECRET), a tile per app, per-user accounts for the whole
team, and board upload so a new diagram reaches the wall without touching a
terminal.
First five minutes
- Create accounts in Settings - one per human, all equal.
- Override a tile URL if an app lives on another host.
- Enable SSO with one shared secret - the README has the recipe.