NetXMS Support Forum

Does 5.0.2 address any possible issues with Neighbors links not being drawn on the map ?  It could possible be a filter issue I'm not sure.  

In order to keep the map visuals under control I set the snmp location to the fiber zone that the switch resides in for instance Zone_1 i then create a corresponding container and map.  Container Zone_1 will have a binding rule to bind all switches that match and an inheritable attribute (Name FiberZone  Value 1) The corresponding map will have the following object filter script.

if ($node->name == "CoreSW") return true;
attributes = $node->customAttributes;
foreach(a : attributes->keys)
{
      if(attributes[a] == 1){
      return true;
      }
}

Prior all switches would populate and draw neighbor links and now any new switch added does not show and if manually added to the map links will still not be drawn.  I am assuming this maybe due to the filtering but I am not sure due to the fact i manually add the device to the map and still nothing is linked.

Thanks  

Previously I was using the following script in the configpoll to use the SNMP sysname

/* Available global variables:
 *  $object - current object, one of 'NetObj' subclasses
 *  $node - current object if it is 'Node' class
 *
 * Expected return value:
 *  none - returned value is ignored
 */
 
if (($node->snmpSysName != "") and ($node->snmpSysName !=null))
{
 RenameObject($node->snmpSysName);
}

I know that RenameObject is deprecated but the new rename() does not seem to work either here is what I tried rename( $node->snmpSysName) is there a better way to do this or can this possibly be implemented in the core functionality or as a server config variable 

What do the yellow DCI mean? they collect data but are yellow.

also script says function "CreateSNMPTransport" and "SNMPGetValue" are deprecated, what are replacements?

Quote from: Filipp Sudanov on May 03, 2024, 12:10:56 PMCurrently you can have "Unbound stuff" container somewhere with this auto-bind script, it will pick objects that are not present in any other containers:

for(p : $object.parents) {
  println(classof(p));
  if ((classof(p) == "Container") and p.id != $container.id) return false;
}
return true;

Thanks Filipp.  

I tried this script however it seems to pickup and autobind all nodes over time, not just those nodes outside the infrastructure services tree with no parent container.  

eg: attached an example of it filtered down to some APs that are already in parent containers and still trigger the autobind event.

5.0.2 is now available. Most likely this will be the last release of the legacy console, we'll push app for the new console in next few days.

These scripts were not converted during migration (for multiple reasons, we'll handle that).

You can convert scripts using "nxscript -5" command.

Copy script from the chart into a file, then run "nxscript -5 file.nxsl", it will output updated script.

5.0.2 in published

Quote from: lpolstor on May 07, 2024, 07:22:30 PMTrying to install Netxms 5.0.1 on Ubuntu 24.04 Server, had some troubles installing

ref  Installation — NetXMS Administrator Guide (4.4.3)  ... ( I assum 5.0.1 version is on it's way..)

Code Select Expand

[color=#404040][size=2][font=SFMono-Regular, Menlo, Monaco, Consolas, Liberation Mono, Courier New, Courier, monospace]wget http://packages.netxms.org/netxms-release-latest.deb
sudo dpkg -i netxms-release-latest.deb
sudo apt-get update[/font][/size][/color]
but when I tried to install netxms-server I got the following list when I tabbed at the end of the prompt:

Code Select Expand

root@testhost:~# apt-get install netxms-
netxms-agent          netxms-base           netxms-dbdrv-sqlite3  netxms-release
Checked  Index of /ubuntu/pool/noble/main/ (netxms.org) and it contains less directories as the same for Jammy, example  Index of /ubuntu/pool/jammy/main/n/ (netxms.org)  has netxms and all the useful deb packages, however the noble one does not.

Regards
Lars Petter

Build for 24.04 is still on the way (there are some changes required and we were busy with other issues), but it will be available in next few days.

They should've been removed when instance DCI was removed from the template. Also "nxdbmgr check" should handle that, if I remember correctly.

Anyway, you can delete them either via API (e.g. using nxshell or custom java app) or by modifying database directly. I'd recommend API.

Something like this (do not run as-is, just an example):
for node in [o for o in session.allObjects if isinstance(o, objects.Node)]:
    c = session.openDataCollectionConfiguration()
    try:
        items_to_remove = []
        for item in c.items:
            if item.name == 'System.CPU.Usage':
                items_to_remove.append(item.id)
       
        for item_id in items_to_remove:
            c.deleteObject(item_id)
       
        if len(items_to_remove) > 0:
            c.commit()
    finally:
        c.close()


I have somehow wound up with a large number of nodes (>1500) that have a DCI that was created by a template dci with instance discovery, but the instance discovery DCI is no longer part of the template. Because of this, if I want to remove these DCI (which I do), I have to go through every node and individually delete each one.
If I go to edit the DCI, I get the popup that says 'This DCI was created through instance discovery' , but there is no instance discovery DCI left on the node.
I am at a loss, it seems there is no way to automate deleting a DCI through API or scripting, and I really don't want to pay someone to go through all of these nodes and sort out which DCI need to be deleted.
Is there a way to bulk delete DCI that I am missing?