The best way to learn about a feature is to use it. We don't have any business services in our environment, so we could use a similar approach as with the network map link indicator feature, where we created fake items and triggers to simulate network issues. We'll create items and triggers that will act as high-level service monitors.

We'll invent two companies, called Banana and Pineapple. Our company will be hosting various services for these two companies:

• A code repository system for Banana
• A warehouse analytics system for Pineapple
• A ticketing system for Banana and Pineapple

Our service tree could look like this:

If everything is green at the top level, we know that all of our customers are happy. If not, we see which customer is having an issue with a system, and we can see which system is affected. The ticketing system going down would affect both customers. And anything below these services—well, that's operational monitoring.

Unfortunately, IT service functionality is not that easy to evaluate without collecting data for a longer period of time; SLA graphs are more interesting when we have data for a few weeks or more. Maybe we could send in data and pretend it's past data. Actually, we can do that. The small but great zabbix_sender tool, which we discussed in Chapter 10, Advanced Item Monitoring, allows us to specify a timestamp for each value. This means that we'll create Zabbix trapper items and push values to those.