Node with question mark on the icon of object browser

Started by tivanni, August 14, 2013, 06:36:37 PM

Previous topic - Next topic

tivanni

Hi!
On netxms, i have a node that cannot be monitored throught icmp ping. On that node, i have a dci that store the info about the status of the node and i manage the alarms depending on the value assumed by this dci.
The only problem is that appear a question mark on the icon of the node in the Object three menu; i think this is caused by the icmp polling disabled .
Is there a way to eliminate the question mark on the icon?
Thank you!

Victor Kirhenshtein

Hi!

The only workaround I've found is to generate alarm for that node with "Normal" state and keep it in outstanding or acknowledged state. I've added a feature request for adding DCI based status calculation (https://www.radensolutions.com/chiliproject/issues/324).

Best regards,
Victor

Tatjana Dubrovica


Dani@M3T

I see this feature is in V1.2.17. But how we can use it. I didn't found a documentation.

tomaskir

Quote from: Dani@M3T on October 20, 2014, 05:11:52 PM
I see this feature is in V1.2.17. But how we can use it. I didn't found a documentation.
DCI > Options > Other Options > Use this DCI for node status calculation

What exactly does it do, I dont know :)

How does status calculation work if this is checked? Can it be checked for multiple DCIs? Etc...

Victor Kirhenshtein

Hi,

as previously noted, you can select one or more DCIs with "Use this DCI for node status calculation". Server expects such DCI to return numeric value in range 0 .. 4 which indicates status (0 =  normal, 1 = warning, 2= minor, 3 = major, 4 = critical). Most critical status of all status DCIs and active alarms will be taken as current node status.

Best regards,
Victor

Dani@M3T

ok. I made my tests to this new feature:

  • node created with no interfaces (for an unmanagable device)
  • in polling properties of this node I deactivated all except 'Disbale data collection'
  • DCI created on this node which checks the status of the switch port (gives back 4 for down and 0 for up) and 'use this DCI for node status calculation' = ON

Now when I switch off the device, I get an event 'SYS_NODE_CRITICAL'. When I switch on the device again, I get a 'SYS_NODE_NORMAL' event.
But for my existing event processing policy I need 'SYS_NODE_DOWN' and 'SYS_NODE_UP' events. So I created a threshold on the DCI (!=0 -> SYS_NODE_DOWN, deactivation event SYS_NODE_UP) and everything works as expected.
I will first use it for unmanageble provider-modems and switches.
Maybe this helps someone.