Zabbix was born as a distributed network monitoring tool with a central web interface where you can manage almost everything. Nowadays, with Zabbix 2.4, the number of possible architectures has been reduced to a single server setup and a Zabbix-proxies distributed setup.

Now, the simplest architecture (which is ready to handle large environments successfully) that you can implement composes of three servers:

To prepare this simple setup for a large environment setting, it's better to use a dedicated server for each one of these components.

This is the simplest setup that can be easily extended and is ready to support a large environment.

The proposed architecture is shown in the following diagram:

This kind of setup can be extended by adding many Zabbix proxies resulting in a proxy-based setup. The proxy-based setup is implemented with one Zabbix server and several proxies: one proxy per branch, data center or, in our case, for each remote network segment you need to monitor.

This configuration is easy to maintain and offers the advantage to have a centralized monitoring solution. This kind of configuration is the right balance between large environment monitoring and complexity.

The Zabbix proxy, like a server, is used to collect data from any number of hosts or devices, acquiring all the metrics requested and acting as a proxy. This means that it can retain this data for an arbitrary period of time, relying on a dedicated database to do so. The proxy doesn't have a frontend and is managed directly from the central server.

All these characteristics make the Zabbix proxy a lightweight tool to deploy and offload some checks from the central server. Our objective is to control and streamline the flow of monitored data across networks, and the Zabbix proxy gives us the possibility to split and segregate items and data on the different networks. The most important feature is that the acquired metrics are stored in its database. Therefore, in case of a network loss, you will not lose them.