How HexitHive works from portal to Raspberry Pi player.
HexitHive treats digital signage like a production system: you model tenants and screens in the portal, publish clean configs, and let hardened Raspberry Pi players sync, cache media locally and keep playing even when the WAN misbehaves.
Step 1: model tenants, users and screens in the portal
Everything starts in the HexitHive Portal. You describe the world you care about: resellers, customers, sites and screens — then give people the right roles to manage them.
Tenant tree that mirrors your real-world structure
The portal lets you create a tenant tree that mirrors how you actually operate: a root organisation, MSP/reseller tenants and end-customer branches or sites. Each screen and playlist belongs to exactly one tenant.
- Root organisation for global admins and MSP teams.
- Reseller / customer tenants for individual brands or organisations.
- Sites or branches as children for each customer.
Roles so people only see what they should
Users are assigned to a tenant with a role, and they only ever see that tenant and its children. No cross-tenant visibility, no “accidental edit in the wrong customer”.
- Root Admin — full system-wide visibility and control.
- Tenant Admin — manages users, screens and playlists for their tenant.
- Operator — works with media and playlists, not structure.
- Viewer — read-only dashboards and lists.
Step 2: upload media and build playlists
Once the structure is in place, you upload media and build playlists that describe what should run on each screen or group of screens.
Media library per tenant
Each tenant has its own media library: videos and images you can reuse across multiple playlists. You see file name, type, size and (for videos) duration.
- Upload once, reuse across multiple playlists and screens.
- Filter by type, upload date or simple search.
- Keep sensitive content isolated to the correct tenant.
Playlists that match how you want content to flow
Playlists are ordered sequences of media items with durations. You can assign a playlist to one screen or many, and change it without touching the players.
- Drag items into the right order (e.g. promo, menu, brand loop).
- Set per-item durations for images, or run videos to completion.
- Assign a playlist to individual screens or entire groups.
Step 3: publish a clean config
When you are happy with playlists and assignments, the portal generates a clean configuration per screen. Think of it as a release: a snapshot of what should be running everywhere.
Accept & Push instead of live-edit chaos
Instead of every tiny playlist edit going live immediately, the portal can treat updates like a release: review changes, then Accept & Push when you are ready.
- Reduce the risk of half-finished playlists going live accidentally.
- Keep a clear mental model of “what is live vs what is still in draft”.
- Give operations a predictable change window for signage content.
Under the hood, this becomes a structured config document per screen — including which playlist should run, which media files are required and which tenant the screen belongs to. Players use this config as the source of truth.
Step 4: players sync configs and media
Raspberry Pi players regularly call home over HTTPS to pull their latest config and any missing media. All connections are outbound-only from the player side.
Secure outbound-only connections
Each player uses a unique device key and only makes outbound HTTPS requests to the HexitHive APIs. You do not need inbound firewall rules or open ports on the LAN.
- Config endpoint such as
/api/player/config.php. - Media download endpoints for video and image assets.
- Optional event and heartbeat endpoints.
Local cache of everything needed to run
Players download a full copy of the playlist JSON and all required media into
a local cache (for example under /var/lib/hexithive/content). Playback
does not depend on a live connection to the portal.
- Predictable sync intervals instead of constant chatter.
- Bandwidth stays local once media is cached.
- WAN problems do not immediately turn into black screens.
Step 5: local playback and offline-safe behaviour
When the WAN behaves, players sync regularly and carry on. When the WAN or portal is unavailable, players keep rendering the last known good playlist from local cache.
Playback loop that only cares about local state
The playback loop reads playlist JSON from the local API and hands each item to the media engine (mpv) in full-screen mode. It never waits on remote APIs to show the next item.
- Videos play to completion or configured duration.
- Images show for configured durations.
- Missing or corrupt files are skipped instead of crashing the loop.
What happens when the internet dies
If the player cannot reach the portal, it keeps using the last successful config and local media. The only visible symptom is in the portal: the screen’s heartbeat goes stale and the status changes to offline or degraded.
- Screens stay on-brand instead of going black.
- NOC teams see which screens are affected via heartbeats.
- When connectivity returns, players quietly resync and carry on.
Step 6: heartbeats, monitoring and maintenance
Heartbeats and local maintenance tools give operations real observability without exposing risky remote-control features to the public internet.
Heartbeats for every screen
Each player posts regular heartbeats back to the portal, including last-seen time and basic health data. The portal turns this into online/offline status for every screen.
- See which screens are online, offline or stale at a glance.
- Use last heartbeat time to spot flaky sites or WAN links.
- Correlate incidents with playlist or config changes.
On-site maintenance tools with a PIN
A maintenance PIN protects sensitive actions on each player. From an on-site maintenance screen, technicians can see IP/DNS state, pairing status and last config sync, and trigger a safe resync or reboot when needed.
This keeps day-to-day operations firmly in the portal, while giving local IT teams just enough access to troubleshoot stubborn sites.
Why the design looks like this
HexitHive borrows ideas from production infrastructure: clear separation of concerns, predictable flows and a bias towards failing safe on the screen in front of people.
Fails safe by design
When something breaks, screens should not go black. Local caches and playback that only depend on local state mean the failure domain is smaller and more predictable.
Predictable and debuggable
Simple, explicit flows — config publish, player sync, local playback, heartbeat — are easier to reason about than opaque “magic cloud” polling. That matters when you are debugging a mall with fifty screens.
Multi-tenant from day one
The tenant model and role system are not bolt-ons. They exist from the first screen and work just as well when you onboard the hundredth customer with hundreds of sites.
Built for operators, not just marketers
Marketers care about campaigns. Operators care about uptime, failure modes and access. HexitHive is built for the latter group, so the former can trust the screens will actually stay on.
Ready to run signage like a production system?
Tell us about your tenants, sites and screens. We will walk through how HexitHive would plug into your world and what the rollout could look like.