In the previous topic where we explained the frontend, we already talked a bit about hosts, triggers, items, and so on. So I am sure you have a lot of questions about what they are. Before we move on with our Zabbix book, I will first explain you a bit about the Zabbix definitions so that you have an idea what everything is.
By now you should have an up and running Zabbix configuration and you should know your way around the frontend. If you have no clue how the frontend works, then go back to previous topic and have a look at Exploring the frontend.
In Zabbix we call hosts, devices that we want to monitor. Of course, those devices need a network connection so that our Zabbix server can talk to them.
Hosts can be added in Zabbix under Configuration | Hosts.
- When we have hosts, it makes sense to group them together based on certain common unique aspects they have, for instance, all Linux or all Windows servers. Groups in Zabbix can contain hosts and templates and are being used to assign access rights to hosts for different user groups. Host groups can be found under Configuration | Host groups.
- Now that we have hosts and host groups, we want to monitor certain things such as memory, CPU load, network interfaces, and so on. This, we call in Zabbix Items. Items can be added on host level or on template level. Preferred way is of course, on template level as we can make use of templates as many times as we want. Items can be found under Configuration | Hosts | Items or Configuration | Templates | Items.
- If we want to monitor hosts, we could create checks for each host or make use of Zabbix templates. Templates are a set of entities such as items, triggers, screens, and so on, put together ready to be applied to one or more hosts. Advantage is that we save time when configuring or making mass changes. Templates can be found under Configuration | Templates.
- Now that we have our items in Zabbix, it makes sense to put certain thresholds that we don't want to pass, for example, CPU load higher than 5 or memory lower than 256 MB. In Zabbix, we make use of triggers to define our thresholds. Triggers are logical expressions that evaluate the data of our items and put an item in a state of ok or problem. Triggers can be found just like items on host level or on template level under Configuration | Hosts | Triggers or Configuration | Templates | Triggers.
- When a trigger changes its state, an event will be generated in Zabbix. Other things in Zabbix that generate events are auto-registration of agents and autodiscovery of network devices. Events can be found under Monitoring | Events.
- Sometimes in Zabbix, we want certain actions to happen based on our events. An action consists of an operation (example: send an email) and a condition (example: when a trigger is in problem state). Actions can be found in the menu under Configuration | Actions.
- Sometimes, sending an email to one person is not enough. It so happens that we want to notify more than one person in a certain sequence. For this, we have escalations. Back to Configuration | Actions under the tab Operations we can add all steps of different escalations we want to follow. Escalations are custom scenarios in an action. For instance, in one action we can send an email and then after 10 minutes send a text message to someone else. There is no limit in the number of escalations steps.
- Media in Zabbix is used to define the way in which we will get our notifications delivered. Remember that media is user dependent and can be found under Profile | Media but also under Administration | Media types. Here we list all media types allowed with their proper configuration.
- Notifications is what we use in Zabbix to notify someone about some event that happened by making use of the selected media channel from the user. Under Administration | Notifications, we can see who was notified at what time by what media.
- Sometimes the options that Zabbix gives us are not enough. For cases like this, Zabbix allows us to extend Zabbix with remote commands. Remote commands are predefined commands that execute automatically on a monitored host under certain conditions. Those can be found under Administration | Scripts.
- When we monitor several network items in Zabbix, it makes sense to put them in a group. This way, it will be easier later to check all data about those items. For this, Zabbix uses applications. When we create items, we can select applications for our items. Applications can be found under Configuration | Hosts | Applications or Configuration | Templates | Applications.
- In Zabbix it is possible to monitor web services. We can build advanced scenarios to check our websites. This we call Web scenarios and can be found under Configuration | Host | Web or Configuration | Template | Web.
- The frontend as we have seen earlier is the web interface from Zabbix.
- When we want to extend our Zabbix monitoring solution, we can do this by making use of the Zabbix API. The API makes use of the JSON Remote Procedure Call (RPC) protocol and can create/update/fetch objects such as hosts, templates, groups, and so on.
- Our environment makes use of a Zabbix server. The Zabbix server is a software process that performs monitoring. It interacts with our agents and proxies, makes calculations, sends notifications and stores all data in a central database.
- The Zabbix agent is the piece of software that we install on our hosts to monitor local resources and applications.
- Nodes are like proxies but with a full server configuration, set up in a hierarchical way. Nodes are deprecated and are removed in Zabbix 2.4, so we can forget about them. Before Zabbix 2.4, we had DM under Administration which is now renamed as Proxies.
The definitions in Zabbix are things you should learn before you really start working with Zabbix as we will be using them throughout the book. It is crucial to know, that when we talk about a host, it can be just any device connected to a network which we want to monitor (example: switches, temperature sensors, gateways, door sensors, printers, and so on). Just as it is important to know what hosts are, it is important to know when we want to monitor something like CPU iowait or the state of a network device. For this, we need to create Items on the host. And that it is best done by creating items on templates and then link a template with one or more hosts. Another thing we have seen is that we have to put certain thresholds on the items that we monitor to get notified about certain problems. This we call Triggers and just like items, we should place them in a template, so that we don't have to create the same trigger over and over.
As always, it's a good idea to check the Zabbix documentation for the latest update about the definitions:
https://www.zabbix.com/documentation/2.4/manual/concepts/definitions.