BIG-IP DNS Generic Host

Hello,

I need to add some services running on 'generic host'. The destination host is reachable (icmp) from CLI (rd0), but after adding it to GTM configuration as 'server' with 'icmp monitor' - it is not getting up (green) but monitor fails with reason 'no reply from big3d: timed out'.

configuration:

gtm server /Common/host-dca {
    datacenter /Common/DCA
    devices {
        host-dca {
            addresses {
                172.25.113.23 { }
            }
        }
    }
    monitor /Common/my_icmp
    product generic-host
    virtual-servers {
        service1 {
            destination 172.25.113.21:0
        }
        service3 {
            destination 172.25.113.23:0
        }
    }
}

gtm monitor gateway-icmp /Common/my_icmp {
    defaults-from /Common/gateway_icmp
    interval 5
    probe-attempts 3
    probe-interval 1
    probe-timeout 5
    timeout 15
}

log:

# host 172.25.113.21
alert gtmd[13178]: 011a5004:1: SNMP_TRAP: Server /Common/host-dca (ip=172.25.113.21) state change blue --> red (No enabled VS available)
alert gtmd[13178]: 011ae0f2:1: Monitor instance /Common/my_icmp 172.25.113.21:0 CHECKING --> DOWN from /Common/dca-dcdns-adc (no reply from big3d: timed out)
alert gtmd[13178]: 011a6006:1: SNMP_TRAP: VS service1 (ip:port=172.25.113.21:0) (Server /Common/host-dca) state change blue --> red ( Monitor /Common/my_icmp : no reply from big3d: timed out)

# host 172.25.113.23
alert gtmd[13178]: 011a5004:1: SNMP_TRAP: Server /Common/host-dca (ip=172.25.113.23) state change blue --> red (No enabled VS available)
alert gtmd[13178]: 011ae0f2:1: Monitor instance /Common/my_icmp 172.25.113.23:0 CHECKING --> DOWN from /Common/dca-dcdns-adc (no reply from big3d: timed out)
alert gtmd[13178]: 011a6006:1: SNMP_TRAP: VS service3 (ip:port=172.25.113.23:0) (Server /Common/host-dca) state change blue --> red ( Monitor /Common/my_icmp : no reply from big3d: timed out)

why is there 'no reply from big3d: timed out'?? the server type is  'generic host' (not 'bigip').

network:

GTM is running in HA (there are two GTM devices). Each of them has one self and one float IP address configured. All in routing domain 0 only. Nothing special.

Troubleshooting

host 172.25.113.21 is currently down, host 172.25.113.23 is reachable:

# ping 172.25.113.21
PING 172.25.113.21 (172.25.113.21) 56(84) bytes of data.
^C
--- 172.25.113.21 ping statistics ---
3 packets transmitted, 0 received, 100% packet loss, time 1999ms

# ping 172.25.113.23
PING 172.25.113.23 (172.25.113.23) 56(84) bytes of data.
64 bytes from 172.25.113.23: icmp_seq=1 ttl=253 time=2.11 ms
64 bytes from 172.25.113.23: icmp_seq=2 ttl=253 time=1.96 ms
64 bytes from 172.25.113.23: icmp_seq=3 ttl=253 time=3.68 ms
^C
--- 172.25.113.23 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 1.964/2.586/3.684/0.780 ms

tcpdump from bigip CLI  (rd0) with above 'ping' tests:

# tcpdump -i0.0:nn -nnp host  172.25.113.21 or 172.25.113.23
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on 0.0:nn, link-type EN10MB (Ethernet), capture size 65535 bytes
11:28:50.368443 IP 172.18.210.20 > 172.25.113.21: ICMP echo request, id 1286, seq 1, length 64 out slot1/tmm0 lis= port=1/0.16 trunk=.itrunk0 flowtype=130 flowid=4000013D1C00 peerid=4000013D1B00 conflags=20E26 inslot=1 inport=18 haunit=0 priority=0
11:28:52.816810 IP 172.18.210.20 > 172.25.113.23: ICMP echo request, id 1321, seq 1, length 64 out slot1/tmm0 lis= port=1/0.16 trunk=.itrunk0 flowtype=130 flowid=4000013CD140 peerid=4000013CD040 conflags=20E26 inslot=1 inport=18 haunit=0 priority=0
11:28:52.818050 IP 172.25.113.23 > 172.18.210.20: ICMP echo reply, id 1321, seq 1, length 64 in slot2/tmm0 lis= port=.itrunk0 trunk= flowtype=0 flowid=0 peerid=0 conflags=0 inslot=7 inport=26 haunit=0 priority=3

Question:

• I thing, the configuration is correct, but server (host-dca) and virtual services (service1 and service3) are going down.
• Why is the reason of virtual service is going down 'no reply from big3d: timed out'? the server type is 'generic host', not 'bigip'
• I can't see monitoring (my_icmp monitor) requests in the tcpdump. why is this not working? I can see only ping request from CLI, but not from monitor 'my_icmp'.

thank you for advice