Individual installations of Little Green Button communicate with each other across a network layer. Early versions had to be within the same subnet. As Little Green Button increased in popularity, it found its way into organisations with more complex network topologies and the single subnet principle became limiting.
To overcome this, engineers at Little Green Button developed a new protocol. It retained all the elegance of lightning fast broadcasts within the local subnet but extended it with the ability to “probe” remote subnets to discover other instances of Little Green Button. This protocol requires little configuration and dynamically maintains connections as machines join and leave the network. This is now known as the “classic” connectivity model. For backwards compatibility, some sites use the original protocol (now known as “island”) where all communication is retained within the local subnet.
Starting in v3.2, Little Green Button engineers added a new option – the “bridge” server. Now, organisations with networks spanning multiple subnets without physical interconnections can use Little Green Button in what is known as the “standard” connectivity model. Little Green Button bridge servers are hosted in a cloud data centre with transparent redundancy being provided by server pairing with automated failover.
When an emergency is raised, the following information about the emergency passes to/from the bridge server.
- A unique identifier
- IP address and station id (formed from MAC address) of the workstation
- Username and identifier of the logon session
- The user-entered name of the location (and telephone number, lat/long and response group id)
- Ids of the target response groups
Similar information relating to any responders also passes across the bridge. Periodically, the count of buttons available within each subnet passes across the bridge too.
Data exchanged with the bridge is encrypted using TLS to prevent in-transit snooping/tampering. Organisations that wish to keep their Little Green Button communications completely restricted within the bounds of their own network can still achieve this with the classic connectivity model.