NetXMS Support Forum

This message indicates that agent was built with libcurl without bearer authentication support. On what platform you run agent and how did you install it?

Could you please share output of command

nddload -c community -n JUNIPER -V path_to/juniper.ndd 10.5.4.2

Replace path_to/juniper.ndd with actual path to juniper.ndd (if installed from packages it likely will be /usr/lib/netxms/ndd) and community with actual community string.

Hello,

currently not, but we definitely plan to move in that direction.

Best regards,
Victor

That's good idea, we will add such banner.

Best regards,
Victor

Remove , at the end of line 51:

   cacOverrun INTEGER,

should become

   cacOverrun INTEGER

Best regards,
Victor

Hi,

it is legacy or new UI? It could be that binding not yet moved to new UI.

Best regards,
Victor

You should put either %a (to use IP address) or #%i (to refer to current node by ID) in "remote host" field.

Best regards,
Victor

Hello,

most likely mysql.ddr cannot find MySQL client libraries. Please check if they are in the PATH.

Best regards,
Victor

Hello!

It is possible that server selected generic driver for this node. Check this in object overview, and if it is the case, try to force use of Juniper driver by adding custom attribute snmp.driver to value JUNIPER. Run cionfiguration poll andf check if VLANs are correctly detected.

Best regards,
Victor

Hi,

this is very strange. I just compared process execution and external parameter handling code in 3.9 and 4.1 and the only difference is that in 4.x agent replaces null characters with spaces in received command output. Theoretically if script returns null character in first line that could cause such difference in behavior.

Best regards,
Victor

So it shows something connected on interface "6", but this interface not connected to anything?

Please run topology poll on that switch manually, and show the output. Also, please show screenshot or export of interfaces table (I'm mostly interested in peer information columns).

Best regards,
Victor

Насколько я понимаю это другое - этот параметр задает, как Juniper будет идентифицировать свои порты в отсылаемых LLDP пакетах.

Hi!

You cannot cancel timer directly from NXSL. However, you can generate special event to cancel timer. For example, in your script you can have something like

PostEvent($object, "CANCEL_TIMER", null, timer);   // Variable "timer" contains timer key

and in EPP you create separate rule that handles event CANCEL_TIMER and the only action is "cancel timer" with %1 as key.

Of course you also have to create event CANCEL_TIMER.

Also, if you need macro string expanded within NXSL, you can use function ExpandString. In your example you can do

timer = ExpandString("nodetimer_%i", $node);

That way you are guaranteed to have exactly same string in NXSL as in other EPP places.

Best regards,
Victor

Hi,

try to use ExternalParameterShellExec instead. Also, how long it take to execute this command? There could have been changes in handling execution timeouts. Agent log on debug level 6 could be helpful as well.

And what version you are using currently?

Best regards,
Victor

Hi,

it could be access control issue - maybe account you are using don't have access to those nodes. Discovered nodes will appear in "Entire Network" tree under appropriate subnet. You can try to login with user "system" and check - this user always has access to all objects, so you should see them or be able to find.

Best regards,
Victor