issue with health monitor instances

The Alias Address and port is used for transparent monitors - where LTM checks an external address through a device such as a proxy or gateway. If you have applied the monitor to the pool then LTM will send that monitor to the ip address and port for that pool member. If you apply the monitor to the node then it will send the monitor only to that node.

You should apply the icmp monitor at the node level as the default node monitor and assign the DNS monitor at the pool level.

Pete, I assigned DNS monitor at the pool level, but in instances i see only one pool member (pool has 2 members) is it ok? How the second member will be checked?

If you have assigned the monitor to the pool then there should be all members in the Instances page. Check at which level you have the monitor assigned and that it isn't at the node level. Feel free to send the output of "tmsh list ltm pool" and "tmsh list ltm node".

My output:

 ltm pool ns {
members {
    10.10.10.17:domain {
        address 10.10.10.17
        monitor udp
        session monitor-enabled
        state up
    }
    10.10.10.18:domain {
        address 10.10.10.17
        session monitor-enabled
        state up
    }
}
monitor dns_health_monitor

ltm node 10.10.10.17 {
address 10.10.10.17
description host_10.10.10.17
monitor none
session monitor-enabled
}
ltm node 10.10.10.18 {
address 10.10.10.18
description host_10.10.10.18
monitor none
session monitor-enabled

You have defined specific member monitor on 10.10.10.17 member.

I think there's only member 10.10.10.18 in your Instances monitor, isn't it ?

Yes, it was enabled udp monitor on member 10.10.10.17, that why I see only 10.10.10.18 in instances. Pete, Thomas thanks for your help.