Zabbix and network fault tolerance

[tore]

I monitor two server centers (in different parts of Europe) with one Zabbix server located in one of the server centers. Since network problems do occur, I get host unreachable triggers going off in the "remote" server center when they are in fact up. How can I get Zabbix to require a host to be down for the last 3 checks (I run these checks every minute) or so, just to rule out network hiccups before sending alerts? I've tried using the count function, but since status isn't changed when the host is down/unreachable this didn't work. Or is there a better way of achieving network fault tolerance for the Zabbix server? I was thinking of using a zabbix poller located on the local network on each server center with a central Zabbix server in one of the server centers, but I guess this will also fail if there's a network hiccup?

--
Tore

[xs-]

Use either the min() or max() trigger functions (refer to docs for more info)
Depending on which numeric value is up/down

[tore]

as I understand I would need a max(180) statement somewhere in {...:status.last(0)}=2 , but I can't figure out how to do this.

--
Tore

[Tenzer]

You will normally replace the "last(0)" part with min/max.

[Alexei]

On a related note, you may want to use ZABBIX proxies for monitoring of remote location after ZABBIX 1.6 is out. This would greatly simplify maintenance and would also guarantee that all the data collected locally will be preserved in case of communication problems.