This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.
You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

Monitoring Port channel Interfaces

I have been digging around for as much information as possible before posting this question and have found one post stating the issue that we see. As of NPM 10.7 it seems that you can monitor the Etherchannel as a whole but not the interfaces that make up the channel (unless you track down those specific interfaces and monitor them that way). Is there any plans in the future that will allow more granular control of the Etherchannels in order to monitor the interfaces associated with the channels?

Thank you

  • I think we'd like to provide deeper visibility into etherchannels.  What would you like to see specifically in NPM?

    By the way, have you seen the etherchannel stuff here?: NTM 2.2 Beta1 - Building a Better Network Diagram Tool

  • I can't answer for cisco, but for Junipers I'd like the interface represented in a hierarchy like this (though rotated through 90 degrees)

    Drawing6.jpg

    [it was faster to use the visio org chart wizard to draw this]

    Here I have an aggregated ethernet (ae) bundle with two xe interfaces and two VLANs on it.

    I normally want to monitor:

    a) the ae budle to see that it is not overloaded

    b) each vlan in the bundle to get per-vlan statistics

    c) the physical interfaces (so I get error counts at the interface level, and can tell if an interface goes down. FCS errors, runts, giants, etc. do not get rolled up to the ae0 stats.)

    Junipers also create per-vlan interfaces below the physical interfaces, personally I'm not interested in monitoring those as they don't show anything interesting (that I can see).

    ===

    note: in non-lagged / etherchannel interfaces I would want to monitor the vlan interfaces below the physical interface because that's where the per-vlan stats are held.

  • specifically I would like to monitor the load balancing of the interfaces within the etherchannels. NTM sounds like a viable solution but do not currently have it in a production environment.