Policy problem/question

Started by xenth, April 10, 2008, 05:50:34 PM

Previous topic - Next topic

xenth

I deleted all the already present policies from the "event processing policy editor" because I want to make
my own.

I also made my own types of events [NETWORK_PRINTER_DOWN] for example, which I gave an alarm message in my
native language.

As source for this policy I gave two printer nodes.

Now the problem:

When the node goes down, the alarm log tells me that the node is down like I configured it, but this error
does NOT go away when the node goes online again, it stays there untill I terminate the error, THEN my other
configured alarm log error shows up saying the node went offline.

I want to get NetXMS to automatically see that the node went back online, and DELETE the error saying that it is offline
so that the error saying that that node went offline stays in the log.

So basically:
1)Printer node goes down -> 2)I get an error saying that it went down [GOOD!] -> 3)The node comes back online -> 4)The error saying it's down stays untill I terminate it -> 5)Now I get the error saying that the node was offline earlier.

I want to make step 4 go automatically , i.e. eliminating the need for human interaction.

How do I do this?

Victor Kirhenshtein

Hello!

This can be done with alarm keys. When you creating alarm from NETWORK_PRINTER_DOWN event, set key for it to something like PRINTER_DOWN_%i (%i will be expanded to node's unique identifier - complete list of possible macros can be found in user's manual). Then, create rule for opposite event, and in alarm column for it, select "Terminate alarm with given key", and enter the same key.

Best regards,
Victor

xenth

#2
Sorry I am not sure what you mean.

Is this what you mean?

1)Make a rule for SYS_NODE_UP
2)At the alarm tab, I check "Terminate alarms" and put SYS_NODE_DOWN_%i there? (stopped using custom events)

Because this doesn't work  :-\

-EDIT
This DOES terminate the SYS_NODE_DOWN error automatically but, how can I get it to terminate the error, as well as show in the alarm log: Node was down ,  or something like this?

xenth

I did it!  For those interested:

Add one policy for the alarm with an unique alarm key
Add another policy for the opposite event to terminate the previous policy with the unique key
Add another policy for the opposite event to display the text: "this node was down blablabla"

Excellent!  ;D