Hello experts 
Really need your help
Zabbix 2.2.5:
Item: agent.ping
[B]trigger: agent.ping.nodata(2m)}=1
If Zabbix 2 can not resolve few hosts fqdn - agent.ping fails for all monitored hosts.
If you try to query agent on the host that is OK during the issue :
[root@ ~]# time `zabbix_get -s host -p 10050 -k agent.ping`
Response time from agent more than 20 sec. Normal response time: ~ real 0m0.304s.
However, ping is OK to the host that is OK. Seems, response time from agent on normal hosts grows during the issue because Zabbix server is too busy..
Zabbix server reports "network issues" for all hosts in server logs, Queue of items to be updates grows.
Also Zabbix fping servers, this item returns correct values during the issue and Zabbix fires issue only for hosts fqdn of which can not be resolved.
Timeout option (for how long we wait for agent) on server side: 25 sec,
Number of pre-forked instances of pollers for unreachable hosts: StartPollersUnreachable=5
Could you please give a small hint, why Zabbix 2 reports false agent.ping alarms if dns cannot be resolved for only few?
Really need your help
Zabbix 2.2.5:
Item: agent.ping
[B]trigger: agent.ping.nodata(2m)}=1
If Zabbix 2 can not resolve few hosts fqdn - agent.ping fails for all monitored hosts.
If you try to query agent on the host that is OK during the issue :
[root@ ~]# time `zabbix_get -s host -p 10050 -k agent.ping`
Response time from agent more than 20 sec. Normal response time: ~ real 0m0.304s.
However, ping is OK to the host that is OK. Seems, response time from agent on normal hosts grows during the issue because Zabbix server is too busy..
Zabbix server reports "network issues" for all hosts in server logs, Queue of items to be updates grows.
Also Zabbix fping servers, this item returns correct values during the issue and Zabbix fires issue only for hosts fqdn of which can not be resolved.
Timeout option (for how long we wait for agent) on server side: 25 sec,
Number of pre-forked instances of pollers for unreachable hosts: StartPollersUnreachable=5
Could you please give a small hint, why Zabbix 2 reports false agent.ping alarms if dns cannot be resolved for only few?