NetXMS Support Forum

Answered part of my own question.  For switches the Index is what I use and it is template-able.  (I just need a 48 port and 24 port template)

That makes it much easier, originally when I noticed the problem was on servers that I am monitoring with agents and they are VMs so the indexes seem off (and varies between one and many interfaces).

Not to bring up an old question, but this was the closest I could find to what I think I am looking for. 

I am trying to create PRTG like bytesin/bytesout per interface monitoring.  I tried to create a template using various variables in the parenthesis, but no luck.  For example, Net.Interface.BytesIn(*) tried things like %{node_id}  But I guess I need to make a DCI for each interface with each interface id.  Is there anyway to automate this?  (Can be a bit tedious when you're doing in/out on several 48 port switches or a 4 blade 6509).

I added a second device and it was placed in the wrong subnet container so now I have 2 10.168.0.0/21.  It is a Cisco router and a WatchGuard (ASA competitor).  I double checked the IPs/subnets on both and they are correct, any thoughts?  I deleted and added back both devices, still no luck.

One device is 10.168.0.1 and the other is 10.168.1.1.

When I right click a node Tools -> connect, it would be wonderful to be able to select RDP, Telnet or SSH.  Additionally, using fancy reverse-ssh tunneling or stunnel or something (in my mind, easier with a linux based server), having it launch the sessions FROM the management server would make my ACLs smaller and more concise (only opening things for the management server and requiring all access through the server).

An alternative, "user-defined" option (maybe configurable from the server's control panel) could be used to launch scripts (where I could define the reverse tunnels or to launch PuTTY from my windows laptop).

If any of this is do-able from the console already and I haven't RTFM'd enough let me know.  I haven't looked into NXSL at all and the I have used the manual to fill in gaps when something wasn't working as expected.

To further elaborate:

When I right click a node and select Layer 2 Topology, the error message is:  "Cannot get layer 2 topology for <node name>: Node has no support for layer 2 topology discovery."

If I right click the same node select Tools -> Info -> CDP, I get the correct CDP listing.  This is for a node that has no LLDP support.

Also, I do have the capabilities pane on the switch which brings me to another (albeit, slight) issue.  Some of Cisco's system descriptions are long and I have to get creative to get to the capabilities pane.  Maybe a scroll bar or description max width with text wrap would be nice.  See attached.  (Despite the cutoff on capabilities, that is the end of the window)

I'm still using the 1.0.10 server, but using the Development console (well, trying the dev console is more accurate, I still have 1.0.10 console installed).  And I did grab the latest one before I posted earlier.  I can manually poll for CDP in NetXMS and see the results.  Are you saying there may be a server error and I should troubleshoot that? (because l2 discovery does use CDP?)

I've been using the development console (I guess it is now Console 1.1.0).  Under Layer-2 discovery I got the error that my device doesn't support it.  Checked it with Info LLDP poll, no output, but CDP does work.  (Some of my Cisco gear is EoL).

It would be great to have the layer 2 topology work from CDP as well as LLDP.

Just to follow up, that was exactly it.  Interesting behavior observed for others who may be searching: I had intermittent up/downs (false positives) as well as sometimes when I would look at the switch it would show me the management servers interfaces and others it would show the switch.

Thanks for taking a look!

Sorry, my attachment was too big, here it is broken in half.

Victor: I removed attachments to prevent unneeded information disclosure.

I am monitoring several nodes in my NetXMS install (on Ubuntu, ver. 1.0.10, sqlite).  I have been getting errors in the log DCI___ Internal: AverageDCPollerQueueSize, DBWriterQueueSize, QueueingTime, PollerQueue and ConfigPollerQueueSize all changing to unsupported (manually clearing buys me about 4 hours before they reappear).  Then a bunch of services go down.  Doing a status poll manually shows it is trying to be done with the wrong IP address (it seems to be using the management IP from a Cisco 6509 (130+ interfaces)).  So I looked at the 6509 node and it had the XMS interfaces only (different from when I initially added it) It had the XMS public IP, internal IP and the NAT Adapter.  So I deleted the 6509 node and all the alarms immediately cleared.  Readded the 6509, populated all the 130+ interfaces (I know because of my 136 emails received seconds after) and all the alams (DCI and servies) went down and unavailable again and the manual poll showed the 6509 address again.

Thoughts?