Object Status is Up, one or more child objects are in Down state.

I'd like to make a correction to my last post. I said that it wasn't resolved in the 10.3 RC3 release, either, but that is not entirely correct.

It seems that it is possible to disable this "feature" in 10.3 and ONLY in 10.3 to show the "correct" status of nodes in maps. I'll elaborate more on why this should be set as the default setting after explaining the feature.

Looking at the table NodeChildStatusParticipation you'll find the following:

As you can see, Orion.UDT.Port status is checked and that means that any status relating to UDT ports - such as 100% used ports on a specific node - will rollup to the overall node status in the map.

You can disable that entity from participating in the child rollup status by removing the check mark or running the following query:

UPDATE NodeChildStatusParticipation SET Enabled=0 WHERE EntityType=’Orion.UDT.Port’

I've tried this myself and can confirm that our maps now show something that is expected, hooray! (Remeber this is only possible in 10.3 RC3)

So, why is it so important that this newly added feature in 10.3 includes disabled rollup setting for UDT by default?

Well to answer this we need to look at UDT and determine what it does really as a module, what is its purpose, what are its functions and what should it alert on by default?

Three things;

1 Is tracking functionality. You can keep an eye on equipment and know where they are located and what they are connected to on your network. Excellent!!!

2 Tracking user activity in your domain by tapping into your ADDSs. Fantastic!!!

3 Inventory for all switches (or anything with a port). This feature allows you to report on port usage; how many ports are used and what are they being used by? You could use UDT to figure out whether you've got more networking equipment that you actually need by connecting nodes from one under utilized network switch to another. In that case, your objective is to use up all the ports on one switch rather than using half of all the ports on two switches. The end result is better use of your equipment and less power consumption. Makes sense, no?

Paying close attention to UDTs third major function, that is - reporting/provisioning of ports, one would expect that if all ports on a switch are being used then that switch is doing what is is supposed to do and everything would be great - for that one switch. Why would that raise an alarm? It makes no sense. And if it does raise an alarm, one would have to ask himself if that is in reality an alarm at all? Do i have to have a minimum of 20% available ports at all times on all monitored nodes? I can't see the logic of alerting on the % used ports in general. It may be relevant and of major importance to someone but then why does that alarm have to create a false positive elsewhere in the system?

Deltona

@All,

On a test deployment with a fresh database, the setting for childstatusparticipation seems to be disabled by default.

It is only on existing databases, or upgraded installations, that this setting is enabled.

fcaron Is there anything we can do to disable the setting from being reverted if or when you ever run configuration wizard?

Is there a database schema that is used to compare settings, which could be modified?