Our environment currently contains 10,600 hosts, 3,500,000 items, 900,000 triggers defined collecting 6800 nvps.
These monitored devices are spread over 800+ geographically remote sites include both on-premise and in the cloud.
We use 12 centrally based Zabbix proxies to perform the collection of the information of all the items.
We are considering implementing a small container based Zabbix proxy at each one of these sites (i.e. 800+) in favor of the central approach; enabling item collection if the WAN links are disrupted.
I would like to know if anyone has configured a large number of Zabbix proxies and if so what lessons were learned?
Were there any performance issues with the Zabbix database (MySQL/PostgreSQL) and the Zabbix server and whether you used Active or Passive Zabbix proxies.
Thanks in advance.
These monitored devices are spread over 800+ geographically remote sites include both on-premise and in the cloud.
We use 12 centrally based Zabbix proxies to perform the collection of the information of all the items.
We are considering implementing a small container based Zabbix proxy at each one of these sites (i.e. 800+) in favor of the central approach; enabling item collection if the WAN links are disrupted.
I would like to know if anyone has configured a large number of Zabbix proxies and if so what lessons were learned?
Were there any performance issues with the Zabbix database (MySQL/PostgreSQL) and the Zabbix server and whether you used Active or Passive Zabbix proxies.
Thanks in advance.
Comment