random hosts dropping off

[milosz]

hi all,

i've been having a problem with windows hosts, all of which show up as available and monitored in the console, and whose agentd logs don't show any errors, not updating their data in zabbix. again: zabbix server is running and can reach the host (telnet to 10051), zabbix_agentd service is running on the client and can reach zabbix server, no errors in the log, nothing in the server debug log. grepping the debug log for the hostname returns nothing. this seems to happen randomly, and after a few days hardly any servers will be updating.

i've tried: rebooting the zabbix server, restarting the zabbix agent service, rebooting the monitored server, and disabling and enabling the monitored servers in the console. the only thing that seems to work is actually removing the monitored server and adding it back in, but a few days later it'll drop off again.

i took a look at the hosts table in the database, and i've found the problem: the disable_until and errors_from fields are populated on the machines that aren't updating. anyone know why (1) this is happening (2) how i can prevent it/further debug it and (3) why i couldn't figure any of this out or fix it through the web console?

thanks,

milosz

[s4ster]

unique host_name

This is a bit later but I was having the same issue after registering several hosts in Zabbix (through a passive proxy) via Ansible. The problem turned out that the host_name happened to not be unique. The existing host with the registered host_name was replaced with the new one, all logs were now associated with the newly registered host, etc.