NetXMS Support Forum

I had discovery running a few days ago:

[17-Nov-2015 07:38:13.433] [DEBUG] AcceptNewNode(172.26.1.3): auto filter, flags=0003
[17-Nov-2015 07:38:13.482] [DEBUG] AcceptNewNode(172.26.1.3): auto filter - bResult=0
[17-Nov-2015 07:38:13.482] [DEBUG] NodePoller: processing node 172.26.1.4/32 in zone 0
[17-Nov-2015 07:38:13.482] [DEBUG] AcceptNewNode(172.26.1.4): auto filter, flags=0003
[17-Nov-2015 07:38:13.532] [DEBUG] AcceptNewNode(172.26.1.4): auto filter - bResult=0
[17-Nov-2015 07:38:13.532] [DEBUG] NodePoller: processing node 172.26.1.5/32 in zone 0
[17-Nov-2015 07:38:13.532] [DEBUG] AcceptNewNode(172.26.1.5): auto filter, flags=0003
[17-Nov-2015 07:38:13.582] [DEBUG] AcceptNewNode(172.26.1.5): auto filter - bResult=0
...
[17-Nov-2015 07:53:20.632] [DEBUG] Finished active discovery check on range 172.26.1.1 - 172.26.1.255


What does flags=0003 and bResult=0 mean? I have Network Discovery configured for Active and Passive and 172.26.1.0/24 is in the Active Discovery Targets, but the above three nodes haven't been added. Also, 172.26.1.0/24 is in zone 11 not zone 0.

Since then I have restarted the netxmsd service and I haven't seen any more active discovery messages. Can I force an active discovery? In the Server Configuration, ActiveNetworkDiscovery = 1 which I assume means 'on'.

Server: 192.168.160.1
Agent: 192.168.160.3
Version: 2.0-RC2

Agent config file:

---

# Log File
LogFile=/var/log/nxagentd
DebugLevel = 3
# IP white list, can contain multiple records separated by comma.
# CIDR notation supported for subnets.
MasterServers=127.0.0.0/8,192.168.160.1/32

RequireAuthentication = no
EnableProxy = yes
EnableSNMPProxy = yes

---

Added agent to server as a new node. Can see details in Overview. Can't poll anything from node though:

---

[2015.11.19 13:43:33] **** Poll request sent to server ****
[2015.11.19 13:43:33] Poll request accepted
[2015.11.19 13:43:33] Starting configuration poll for node monitoring.xxx.local
[2015.11.19 13:43:33] Checking node's capabilities...
[2015.11.19 13:43:33] Capability check finished
[2015.11.19 13:43:33] Checking interface configuration...
[2015.11.19 13:43:39] Unable to get interface list from node
[2015.11.19 13:43:43] Interface configuration check finished
[2015.11.19 13:43:43] Checking node name
[2015.11.19 13:43:43] Node name is OK
[2015.11.19 13:43:43] Reading list of installed software packages
[2015.11.19 13:43:43] Unable to get information about installed software packages
[2015.11.19 13:43:54] Finished configuration poll for node monitoring.xxx.local
[2015.11.19 13:43:54] Node configuration was not changed after poll
[2015.11.19 13:43:54] **** Poll completed successfully ****

---

I can use nxget to fetch info:

---

# nxget 192.168.160.3 'Net.Interface.Description(eth0)'
eth0

---

Turn up debugging on agent to 5, restart agent, try again:

---

[19-Nov-2015 13:42:59.183] [INFO ] NetXMS Agent started
[19-Nov-2015 13:43:02.576] [DEBUG] Incoming connection from 192.168.160.1
[19-Nov-2015 13:43:02.576] [DEBUG] Connection from 192.168.160.1 accepted
[19-Nov-2015 13:43:02.659] [DEBUG] Incoming connection from 192.168.160.3
[19-Nov-2015 13:43:03.091] [DEBUG] Connection from 192.168.160.3 rejected
[19-Nov-2015 13:43:03.091] [DEBUG] [session:0] Established proxy connection to 192.168.160.3:4700
[19-Nov-2015 13:43:03.091] [DEBUG] [session:0] CommSession::disconnect()
...repeats...
[19-Nov-2015 13:43:58.190] [DEBUG] [session:0] Session with 192.168.160.1 closed

---

...so it's using itself as a proxy to talk to itself, but not from 127.0.0.1, from its interface that faces the server.

Is this the expected behaviour? The zone that it's in *did* have this server set up as the default proxy for the zone, but I put it back after it stopped working [I guess I know why now]. Checked the node's properties, it is not configured to use itself to poll itself.

Right now [2.0rc2] the management console window title is "NetXMS Management Console". I think it would be sensible to add [either by default or otherwise] the hostname of the server you're connected to and the user you're logged in as [username@hostname].

Another that might be useful is the number of unacked alarms.