Zabbix as a CMDB solution

[fdcsantos]

I'm planning to migrate from Nagios to Zabbix. However, one of the most important feature to me is to monitor applications. I read on its features that it's possible, but what I've been expecting of it is to group all the components that belongs to an specific application (database, network connection, server and so on) into a logical group, which means that if every monitored piece of that application is ok, means that my application is ok too. To make Zabbix focus on business as noticed into ITIL framework.

Just an example what I'm looking for is to get one specific interface of a switch, for example, and group it as part of an application group to be more precisely into our monitoring purpose? The objective is to dismiss the possibility of false positives monitoring alerts. I mean, if there is a switch that connects various servers, if an interface goes down, it could or not belongs to that application.

Is there any possibility to do something like this, or what is the solution of community to aim may objective?

[zabbix_zen]

Oi Flávia.

if every monitored piece of that application is ok, means that my application is ok too

Zabbix allows this logical grouping you expressed need for,

For instance, we've grouped our hosts by clients and are using a World Map with their geolocations, if a Trigger goes ON in one of those groups, the corresponding location will go red (useless, but with an 'Wow preety effect')

Nagios and Zabbix have different philosophies,
though Nagios looks for each 'service' health, Zabbix cares about Triggers.

if there is a switch that connects various servers, if an interface goes down, it could or not belongs to that application.

You have Trigger's level dependency and not host level as in Nagios, so if interface eth1 goes down and only Applications A and C are dependent on it, (and you configured some Trigger 'App A' and 'App C' with a parent dependency to that interface one.
Trigger 'App A' <-- Trigger 'interface eth1'
Trigger 'App C' <------------|

In this case if all 3 services are down, you'll only get a notification about 'eth 1'.)
Your Triggers for application B won't obviously care.

[fdcsantos]

First of all. Thanks for your return.

Well, your tip helps a lot. However, I found three points that should be considered into your solution. Please be free to tell me if I did something wrong.

1) When a trigger is created, it needs to belong to a host. Or maybe is only shown that way.
2) When an alert for my trigger called "Amazoncred teste" shows up on my Dashboard, it only tells there is a problem, but no detail is shown (in this case I configured port 22 to alert when UP). See Dashboard snap.
3) To see the conditions of a trigger alert, I need to see its details into Configuration -> Hosts -> Triggers. See Triggers snap.

[zabbix_zen]

1) Correct.
You can think of Zabbix as the following flow,
Hosts (which can be grouped into an Host Group and get Templates attached to them) -> Items -> Triggers -> Actions (which can trigger notifications for the users with Read access to the wanted Host Group(s) via the set upped Media Types)

2)
The Trigger name should be something that gives a description of what just went activated. You can make use of macros like {ITEM.LASTVALUE} in it's name.
Zabbix comes with built-in Templates, I take it you took advantage of net.tcp.service[ssh].last(0)}=0 for your SSH port case.

3)
Correct. Probably the Trigger conditions were interestedly left out of the Dashboard to avoid clutter.


PS: Thought you can find help from our Portuguese/Brasilian community in its subforum, considering English is obviously not a problem for you, if you want to target a wider community I suggest you switch your Zabbix install from Portuguese to English(both helpful for you stating questions and for other users replying).