well, last night I got 45 false alarms, that cleared themselves in a minute - all for devices being 'down' - however, they never went down, and Cacti, Smokeping and Hyperic all showed no problems.
Honestly, Zenoss will only say that a host is down if it fails to respond to pings. It could be a lot more sensitive than Cacti, SmokePing and Hyperic. If you look at the settings for the status monitor, you'll see that the default is to make two ping attempts and to wait 1.5 seconds for each one to come back.
You could verify what is actually happening quite easily by running a tcpdump all night long filtering for ICMP packets only. Then you can go back to a specific timestamp when you get these alleged false positives and see the ICMP echo requests go out, then see whether or not the echo replies came back withing 1.5 seconds.
If the replies do come back, but are delayed you could simply adjust the status monitor's configuration to allow for more latency. If they don't come back at all you may need to raise the number of tries.
_______________________________________________ zenoss-users mailing list [email protected] http://lists.zenoss.org/mailman/listinfo/zenoss-users
