Loading ...
Sorry, an error occurred while loading the content.

How to minimize Nagios alert notification timing ?

Expand Messages
  • tux.ambarish
    I have configured Nagios server for monitoring our 169 Hosts and 411 services. But my Nagios server detect alert after more than 35/40 minutes. Which
    Message 1 of 3 , Mar 15, 2010
    • 0 Attachment
      I have configured Nagios server for monitoring our 169 Hosts and 411 services.
      But my Nagios server detect alert after more than 35/40 minutes. Which parameter I have to change in nagios configuration files so my alerts gets detected immediately.

      I have enabled Active Checks for Host and Services.

      I have check with command nagiosstat ( as well as in web console ) , i found

      Active Service check Latency is: 664.64 sec
      Active Host check Latency is: 662.32 sec

      I want to know both above mention factor are make delay in alert notification, and if Yes how to reduce that
    • Scott
      ... I have a feeling it s in hosts.cfg, but don t have time to go through my config files in depth at this moment. Take a look at
      Message 2 of 3 , Mar 15, 2010
      • 0 Attachment
        On Mon, Mar 15, 2010 at 12:49:53PM -0000, tux.ambarish wrote:
        > I have configured Nagios server for monitoring our 169 Hosts and 411 services.
        > But my Nagios server detect alert after more than 35/40 minutes. Which parameter I have to change in nagios configuration files so my alerts gets detected immediately.
        >
        > I have enabled Active Checks for Host and Services.

        I have a feeling it's in hosts.cfg, but don't have time to go through my
        config files in depth at this moment.

        Take a look at

        http://www.maxsworld.org/index.php/how-tos/nagios

        and see if Max mentions it somewhere in there.
        Sorry to be of relatively little help here.

        --
        Scott Robbins
        PGP keyID EB3467D6
        ( 1B48 077D 66F6 9DB0 FDC2 A409 FA54 EB34 67D6 )
        gpg --keyserver pgp.mit.edu --recv-keys EB3467D6

        Angel: I knew this was gonna happen.
        Buffy: What? What do you think is happening?
        Angel: You're 16 years old, I'm 241.
        Buffy: I've done the math.
        Angel: You don't know what you're doing. You don't know what you
        want.
        Buffy: Oh, no. I think I do. I want out of this conversation.
      • Sheryl
        ... It s never going to be immediate if you just use polling. You can reduce the time. The default IIRC is 10 minutes, and then it waits until the service
        Message 3 of 3 , Mar 15, 2010
        • 0 Attachment
          > I have configured Nagios server for monitoring our 169 Hosts and 411
          > services.
          > But my Nagios server detect alert after more than 35/40 minutes. Which
          > parameter I have to change in nagios configuration files so my alerts
          > gets detected immediately.

          It's never going to be immediate if you just use polling. You can reduce
          the time. The default IIRC is 10 minutes, and then it waits until the
          service fails a few times before it alerts. Think about it, if there's a
          temporary blip in the network, do you want all the alarms to go off?

          > I have enabled Active Checks for Host and Services.
          >
          > I have check with command nagiosstat ( as well as in web console ) , i
          > found
          >
          > Active Service check Latency is: 664.64 sec
          > Active Host check Latency is: 662.32 sec
          >
          > I want to know both above mention factor are make delay in alert
          > notification, and if Yes how to reduce that

          I don't know where those figures come from. I think the default time
          between checks is set in objects/templates.cfg in the newer versions of
          nagios (I'm heading out the door so no time to look it up right now).
          Check the docs, you can override that on a service-by-service basis. I
          think there's also a way to tell it to alarm on the first failure instead
          of waiting.

          But think about how much of that you really want to do. Depending on the
          number of machines and services you monitor you could reach a point of
          diminishing returns by overloading your server or network.

          Another way to look at it is to run something on the server you're
          monitoring to check what you're interested in on each server and then have
          the script send to nagios via nsca. Nagios will alarm if it's expecting,
          say, a status every 5 minutes and you haven't seen one in 7 minutes.

          If you're monitoring microsoft systems you probably should have traps set
          up to send to nagios and you may be able to do that with Net-SNMP on *NIX
          boxes. I just haven't tried it.

          So like most things in *NIX, there are many ways to accomplish the task.
          It all depends upon your environment.

          Good Luck,

          Sheryl
        Your message has been successfully submitted and would be delivered to recipients shortly.