Zabbix Agent and Agent2

[stevefxp]

Hi all,

Can these two peacefully coexist together on one system, or is it one or the other? I am finding there are things the Agent can report on that Agent2 does not.

Thanks,
Steve

[cyber]

On different ports, yes...
I think agent2 can do everyting classic one does, plus it has more built-ins...
even docs say so .. https://www.zabbix.com/documentation...ent_comparison

UNIX: All metrics supported by Zabbix agent.
Additionally, the agent 2 provides Zabbix-native monitoring solution for: Docker, Memcached, MySQL, PostgreSQL, Redis, systemd, and other monitoring targets - see a full list of agent 2 specific items.

Windows: All metrics supported by Zabbix agent, and also net.tcp.service* checks of HTTPS, LDAP.
Additionally, the agent 2 provides Zabbix-native monitoring solution for: PostgreSQL, Redis.

[stevefxp]

I have my Zabbix server configured as in the firat screenshot. Is it normal then for the server to show an unknown status, as in the second screenshot? I am trying to get a baseline of agents to use on a normal 3 tier Linux server.

[stevefxp]

What I found is that I do not need to specify the interface information, when using the active agents. That is quite interesting...

[MRedbourne]

Generally, I would use Agent2. I wouldn't bother using the previous version of agentd unless there is some extreme niche use not documented by Zabbix.

As for information and interfaces when using Agent2 in active mode: Yes, this is true. That's because the system is pushing its FQDN when sending the data to Zabbix which is matched against the list of hosts being monitored. Agent2 also (as of 7.0) has full support for active command execution as well. If you're below 7.0 you will need to specify the interface (<IP/FQDN>:10050) for command execution to work.