Quick Start

This guide gets you a running Omniglass stack for evaluation. It is not a production deployment — defaults prioritize getting started quickly. See Going to Production for the checklist before deploying this to monitor real infrastructure.

Prerequisites

An amd64 host with:

2+ vCPUs and 4+ GB RAM
20+ GB free disk
Docker Engine and Docker Compose v2
A browser that can reach the host on port 80

Linux, Windows (WSL/Docker Desktop), or Intel Mac all work. Apple Silicon needs amd64 emulation (e.g. UTM).

Verify Docker:

docker --version
docker compose version

1. Download and start

mkdir omniglass && cd omniglass
curl -LO https://github.com/hyperscaleav/omniglass/releases/latest/download/docker-compose.yml
docker compose --profile server up -d

No configuration file is required for local evaluation — the defaults work out of the box.

First-time startup pulls images and runs database migrations. Allow a minute or two for everything to come up:

docker compose ps

2. Log in

Open http://<your-host-ip>/ in a browser (or http://localhost/ if you’re running Docker on the same machine).

Username: Admin
Password: omniglass

Change the password after first login.

3. Explore

Omniglass loads with example systems, locations, device types, and hosts so you can see the product immediately. Try the Systems Explorer under Monitoring to see Omniglass-specific views, or explore the standard Zabbix dashboards to see what it adds to.

When you’re ready to add your own devices, read Systems & Locations.

Optional: edge deployments

An edge deployment collects from AV devices at a remote site and forwards data back to your Omniglass host. Use edge deployments when devices live on isolated networks, at other buildings, or anywhere your Omniglass host can’t reach directly.

Each edge deployment runs a proxy + local agent at the remote site. Omniglass must know about it — both the registration on the host, and the registration in the Omniglass UI.

1. Register the edge in Omniglass

Before you start anything at the remote site, register the proxy in the Omniglass UI:

Administration → Proxies → Create proxy
Proxy name: a unique name for this edge (e.g. site-a)
Mode: Active
Save

Remember the name you used — it becomes the EDGE_HOSTNAME for the deployment.

2. Deploy the edge

On the remote host:

mkdir omniglass && cd omniglass
curl -LO https://github.com/hyperscaleav/omniglass/releases/latest/download/docker-compose.yml

Create a .env file next to it. Edge deployments require both variables:

SERVER_ADDRESS=your-omniglass-host
EDGE_HOSTNAME=site-a

SERVER_ADDRESS — hostname or IP of your Omniglass host, as reachable from this edge.
EDGE_HOSTNAME — a unique name for this edge (e.g. site-a, building-b). Must match the proxy you registered in Omniglass.

Start the edge:

docker compose --profile edge up -d

Within a minute the proxy will appear as connected in Administration → Proxies on your Omniglass host.

Managing the stack

Stop (keeps your data):

docker compose --profile server down

Start again:

docker compose --profile server up -d

Upgrade to a new release:

curl -LO https://github.com/hyperscaleav/omniglass/releases/latest/download/docker-compose.yml
docker compose --profile server pull
docker compose --profile server up -d

Always check the release notes for breaking changes before upgrading.

Start over from scratch:

docker compose --profile server down --volumes

Ready to deploy Omniglass for real? Read Going to Production for the checklist.