I recently had a Zabbix failure where the Zabbix server network interface was down for several hours, and the result was that hundreds of triggers went into alarm. The alarming triggers of course couldn't send alerts, but when the interface was brought back online, Zabbix sent out hundreds of recovery alerts.
Shouldn't the Zabbix server know that there are no active interfaces? This seems like a defect/bug, but I'm also wondering if I'm just missing something in my config. I thought about some sort of global dependency, but modifying every template/trigger to include a dependency on Zabbix server eth0 status seems like a brute force method.
Shouldn't the Zabbix server know that there are no active interfaces? This seems like a defect/bug, but I'm also wondering if I'm just missing something in my config. I thought about some sort of global dependency, but modifying every template/trigger to include a dependency on Zabbix server eth0 status seems like a brute force method.