After a few minutes have passed, check the Monitoring | Discovery section:
All the devices that respond to the ping in the configured range will be listed here. If a device is already monitored as a host in Zabbix, it will be listed in the Monitored host column. We will also see for how long the host is known to be up, and, in the ICMP ping column, we also see the time the service is running. In older versions, this column would just be marked green and we would have had to move the mouse over the column to see how long the status was running. It can happen that only one host is listed, as already monitored here. How come? Hosts are recognized here by their IP addresses, so we used 127.0.0.1. The address by which it was discovered differs, so it's not really considered to be the same host or device.
Now, follow these steps:
- Navigate back to Configuration | Discovery and click on A test discovery in the NAME column. Click on New in the Checks block and choose a service that is accessible and would be easy to control on these hosts – perhaps SMTP again. Click on Add in the Checks section, and then click on New there again. This time, choose a service that is not present on any host in the configured range—FTP might be a good choice. Then, click on Add in this block again:
- Finally, click on Update. After a couple of minutes, visit Monitoring | Discovery:
SMTP has appeared, which is great. But why is there no FTP column? Could this view be limited to two services? It's not limited to a specific number of services, but a service that is not discovered on any of the hosts does not show up at all at this time. If a service was initially discovered on a number of systems but not on others, the column would be shown and the systems where the service was not discovered would get a gray cell.
Let's break something now—bring down the SMTP service on one of the hosts, and wait for a couple of minutes. The SMTP cell for that host should turn red, and the popup should start tracking downtime for that service now. If all services on a host went down, the host itself would be considered as down, and that would be reflected in the Uptime/Downtime column:
