Nodes with Problems view & Active Alerts

Ran into an issue where a client has ISDN backup circuits. Of course we want them monitored, but wanted to be alerted when they are up. Modified the Interface alert to not trigger when these are down, have a separate ISDN alert for when they are up, though the node still shows up in the Nodes with Problems view. Is there anyway to exclude this behavior?

Oh and it would be great if we could clear alerts from the web interface after we've ack'd them ! (next service pack or v10 please...)

Find more posts tagged with

Accepted answers

All comments

FormerMember

That thread shows how to use a custom property to use as a filter (in this case we used CustomerName) for the view. Sledgehammer as the fly swatter. What I would like to see is this functionality pushed to the node/interface management level so that we can be more granular on alerts.

FormerMember

What I would like to see is this functionality pushed to the node/interface management level so that we can be more granular on alerts.

Can you explain what you mean in more detail?

FormerMember

and it would be great if we could clear alerts from the web interface after we've ack'd them

In the alerts view, you can hide acknowledged alerts. What do you mean when you say "Clear"? What else is missing?

FormerMember

I apologize for bringing up two things in one thread.

Nodes with Problems - In this case, the ISDN bearer channels will always stay down and therefor NPM see's that as a problem. If I could choose what state is a problem for the node/interface, then I can populate that view with things that are real problems. Example of monitoring user interfaces and excluding them from certain alerting would be good (up/down I don't care about but massive errors I do).

Clearing Alerts from the Advanced Alert View - you can acknowledge alerts from Alert Manager on the poller and you also have a box you can clear the alert from the view. In the web interface you can acknowledge alerts, but you don't have that box to clear them from the view. I don't want to have to log into the poller to clear an alert. What prompted this was that we started needing hourly and daily 'reminder' alerts for certain functions of our NOC (backup checks, log reviewing, to show when tasks are happening, etc). These were based off of basic conditions that would never change (check every hour if node X is up, trigger an alert). Because I don't want 10 engineers trying to RDP to the pollers to clear an alert, I created file monitor in APM with a Windows scheduled task to create and delete a file at given times. This triggers and clears the appropriate alert. I thought this was pretty involved for such a simple need, though if you have an alternative way, I'm all for it, I have tons more to create for each of the operations groups.

FormerMember

Ok. Got it. Many thanks.