What's the difference between 'zabbix agent' and 'zabbix agent (active)'?

[mdeavila]

What the difference between a regular agent check and an active check?

i have not been able to track down any documentation on this topic ...

thanks,

Miguel

[Wolfgang]

My understanding is the following:

a regular agent check is a built in check. That is, a check that zabbix supports out of the box.

an active check is a check, that requires a a script or external command to be defined in the agent configuration file and that is executed via the agent. the result of that script is passed back to the zabbix_sever.

for example:
you want to get the status of an raid array of an server. The raid tools allow a status query for the raid status via a specific command line interface tool. Of course, zabbix won't have native support for this particular raid controller, so you can define an active check that executes a specific query-command via the CLI to get the status.

[Alexei]

Ignore all what Wolfgang said! It is absolutely not correct..

A regular check is initiated by ZABBIX server, it periodically sends requests to an agent to get latest info. The agent is passive, it just processes requests sent by server.

An active check works the following way. ZABBIX agent connects to ZABBIX server to get list of all checks for a host. Then it periodically sends required information to ZABBIX server. Note that ZABBIX server does not initiate anything. ZABBIX agent does all active work. This doesn't require polling on server side, thus it significantly (1.5x-2x) improve performance of ZABBIX server.

[Wolfgang]

Upps ;-)

@Alexi
Thanks for clarification.

[Alexei]

No worries!

[mconigliaro]

so is it safe to assume that active checks should be used whenever possible? are there any disadvantages?

[Alexei]

I would recommend using at least one 'normal' check (for example, TCP or ICMP ping) per host. Otherwise you'll not be able to monitor their availability.

[bbrendon]

active agents behind firewalls

Is there going to be a way to configure the host configuration screen so you can have agents that are behind firewalls and have proper availability information? Right now all my agents behind a firewall report "Not Available" under the Availibility column. I was thinking Availibility should be "Not available" if the server hasnt heard from the agent is say, 5 or 10 minutes.

In addition, in the error column I see "gethostbyname() failed [Unknown host]", because resolution for these systems also doesn't work where the server is located. I'm not sure how to solve this, but it doesn't seem to be a problem, just an annoying error.

I'm using alpha12 (not beta1 because of the bugs).

[ruckus37]

In addition, in the error column I see "gethostbyname() failed [Unknown host]", because resolution for these systems also doesn't work where the server is located. I'm not sure how to solve this, but it doesn't seem to be a problem, just an annoying error.

I'm using alpha12 (not beta1 because of the bugs).

Just add them to the host file...

[slashx]

active details

with active checks, how often to the results get sent back to the server and how many daemons, if not all, process the active checks on the client.

sx

[James Wells]

Greetings,

with active checks, how often to the results get sent back to the server and how many daemons, if not all, process the active checks on the client.

The results get sent back to the server based the periodicity / delay you choose for the item itself. Only a single thread will process the active check on any given time slice. This means that if you have a delay of 60 seconds, only a single thread / agent every 60 seconds will perform the check and then send the results back to the server.

[slashx]

active and passive checks

does enabling active checks on the endpoint automatically disable passive checks by the server for any individual endpoint.

sx

[James Wells]

does enabling active checks on the endpoint automatically disable passive checks by the server for any individual endpoint.

Not at all, you can freely mix the two without issue.

[slashx]

disable passive checks

how do u disable passive checks.

[James Wells]

how do u disable passive checks.

Only way to really disable passive checks to to disable them in the UI, or just make all checks active. Please note that this will not truly disable them as the agents still periodically gather certain information. Additionally, even if you convert all of your checks to active checks, the agents themselves will still listen for non-active checks.