4 Replies Latest reply: Feb 23, 2012 9:00 AM by bobross RSS

Separate Node alerts

grahampriley
Currently Being Moderated

Hi there,

I'm sure this is possible but I'm not sure how.

We have the standard Node 'reboot' and 'down' alerts configured but we want to be able to differentiate between servers and network devices like routers / switches, etc.  It seems at the moment these are both classed as 'Nodes'.  Pretty much like separating the NPM and APM alerts.

Any ideas?

Many thanks,

Graham

  • Re: Separate Node alerts
    nrms
    Currently Being Moderated

    Am I right in thinking you want a different Subject / Body depending on the device type?

    If that is what you're after, create a Custom Property called something like "Device Type" and populate that on all your servers with "Server" all your switches with "Switches", etc. You can then grab that Custom Property's contents when you create your alert by inserting a variable.

    • Re: Separate Node alerts
      grahampriley
      Currently Being Moderated

      Well we just want the Networking folks to be alerted about the Routers, Switches, etc and the Server guys for the servers.  At the moment, both are classed as 'Nodes' so everyone is alerted about everything!!

      • Re: Separate Node alerts
        aLTeReGo
        Currently Being Moderated

        Custom properties are certainly one way but a simpler method might be to leverage the vendor in the alert trigger so they only apply to the nodes that are applicable based on vendor. Below is an example for Windows and Linux servers. 

        • Re: Separate Node alerts
          bobross
          Currently Being Moderated

          We've found that it works better for us to use Custom Properties instead of Custom Alerts.  We were originally going with custom alerts, but as more different things got added we found that there were just too many places where changes needed to be made.

          With Custom Properties we were able to get it down to a handful of alerts.  We have a few Custom Properties that we use for triage/routing:

          AdminGroup (Used by triage team for escalating to the proper group)

          AdminEmail (Email addresses for Node,Volume,Interface Alerts - Alerts delivered directly to addresses listed)

          AppAdminEmail (Email addresses for Applications - Alerts delivered directly to addresses listed)

          Going with Custom Properties overall lets us have a bit more control over the alert structure by allowing us to set the Warning and Critical thresholds for various elements without the need for creating a bunch of custom alerts or alerts with insane boolean logic.

More Like This

  • Retrieving data ...