Infrastructure

Build with granular control for users and workflows, customized connections, flexibility, scalability, and security
Platform
Scalability

Easily scale from a few
screens to thousands at
hundreds of locations
worldwide. The 22Miles
platform can grow with you
from a single-core virtual
machine to a server cluster.

High
Performance

Multi-tier content caching
schema minimizes network
traffic. Push messaging
reduces communication
delay by 90% and reduces
the cost of large-scale
deployment by over 30%.

High
Stability

99% average server uptime
and 99.99% player uptime
across all customers in the
past 15 years.

Highly
Secured

Modern zero-trust architecture
approach ensures security
controls with strong user
authentication with SSO and
MFA support, device security,
user-friendly permissioning
roles, and secure
application controls.

We take your Connection Seriously.

Customize the Connection

Regardless of industry, type of application, or scale of deployment, the 22Miles platform can be
easily adjusted to your requirements. From servers to software, the platform is built to adapt.
On-Premise Enterprise Server

enterprise server, server, digital signage for enterprise

Standard Cloud Hosting

cloud hosting, CMS software, signage solutions

Engineering Documents

engineering documents, servers, signage servers

Enterprise Cloud, Dedicated Cloud, and On-prem deployment options

Global Scalability FAQ

Scalability

How many devices can connect to a single server?

  • Signage devices AWS M5.xlarge windows can handle 500 devices, signage, or wayfinding. M5.4xlarge windows can manage 1500+ devices. In a large enterprise deployment, we recommend the following servers:
    – Main Server – users, projects, database, etc.
    For global distribution, a client may consider multiple media servers at different locations, like the US, EU, Asia, etc. This server hosts the media files with a higher traffic load.
    – Ping Server – monitoring server. We usually separate this server for large deployment.
    – Statistics server – for wayfinding/touchscreen statistics. We also have a proof of play server, but typically only ad-related customers order this server. This server is the top load server.
    Note: each media player, including signage, wayfinding, and scheduling panels, will cache all templates and media files on local storage – assume with windows, brightsign, android, WebOS, Tizen, or RPI 4 OS.
    Any server failure won’t impact the devices’ performance.
  • Wayfinding devices Doesn’t matter which type of application
  • Scheduling panels Same. But this one would require a dedicated API server for 200+ devices. We need to subscribe to 365 changes vs. pull for every panel inquiry. Note: if the API server fails, it will impact all booking panels.
  • Users with mobile devices Start with M5.large should be ok. Mobile should be a separate server. Please note mobile will refresh content upon each visit – the mobile browsers handle caching. If the client has a huge number of concurrent visits, we need to add more mobile servers – not just upgrade the server as traffic/bandwidth could also be a bottleneck. If the mobile server fails, it will impact all mobile access.

What is the response time for each device connected to a single server?

This doesn’t apply. We have multiple communications for different purposes: check update has two options: always-on connection vs. pull; send status to ping server; downloading updates; etc. Any server

What is the recommended server design for 1000 devices, 2000 devices, 3000 devices (etc.)

Main and Ping servers: Minimal M5.2xlarge for 1000, 4x large for 2000, 8xlarge for 3000,
Media server: depends on file and locations. Per continent/region should be ok. This doesn’t require many system resources except storage.
Statistics server: it depends on how many touchscreen devices and usage. Can start with M5.2xlarge

What is the recommended design for a global deployment?

Regional servers / Single server for all sites/all devices? [see above]

Redundancy

What is the recommended design configuration for a 22Miles OnPrem (or private cloud) deployment for touchless/seamless failover?

As said, the servers for devices may not need HA. Any of the servers won’t affect players. If the primary server is down, users won’t be able to make updates, but all schedules have been sent to each device and cached.
O365 API server can be on HA. Not sure about failover – to support this, we have to subscribe to 365 updates on both servers, which generally should be avoided.
A mobile server can be on HA or any failover infrastructure. This server doesn’t have a database and can be easily replicated – when the interface updates, we can publish the content to multiple servers.
If on 22M private cloud

Reliability

What is the recommended OnPrem (or private cloud) server design for 99.999% uptime?

On prem server is designed to 99.999% uptime. We rarely have tickets for on-prem servers. The challenge mainly comes from the initial installation of anti-virus software. These should be all one time to deal with. For the private cloud, the challenge is data center uptime which is not controlled by 22MILES. We use two providers: AWS and iONOS. The former’s benefit is cost, and they don’t meter traffic.