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.

Traffic reported is more than bandwidth available

Hi all,

First time thwack poster here. I was wondering something. We have a remote office (offshore) that is limited on 1 Mbps bandwidth, however the traffic reported have go up to 4Mbps. Anyone mind telling how is this possible? Seems like the data not accurate enough if I continue using it.

NetFlow traffic.JPG

Appreciate if someone able to advise, thanks!

Regards,

LouiE

  • Looks like you are using Riverbed appliances which could be compressing your data. If your NetFlow stats are associated with a port in front of the Riverbed appliance then the amount of traffic reported can be greater than you connection size. The Riverbed device will compress this data to get it across the link

    Darragh

  • So if I were to get the correct traffic utilization for this particular router, all I have to do is exclude the traffic that going and out of Riverbed? Is there any other elegance way to sort this out?

  • Take a look at this, if the diagram matches up with your configuration then this is an option

    http://www.solarwinds.com/documentation/Orion/docs/OrionAndRiverbedWANOptimization.pdf

    Another option is to export NetFlow directly from the Riverbed device but I would try the suggestions at the link above first.

  • FormerMember
    0 FormerMember in reply to louie

    You're seeing the optimised stream so it should be no bigger than the bandwidth of the link serving it. As mentioned above, if you were looking at the pre-optimization traffic you would expect to see much bigger flow rates that the link can actually support. However, if looking at pre-optimization you would not expect to see port 7810. Port 7810 is normally used for Out of Path or Fixed Target optimisation rules. This means that optimised traffic could be using the Riverbeds in-path interfaces or the Primary interface or both. You can set up all these interfaces to send netflow data so get a better picture of which Riverbeds are at the other end of the conversation. That might show up some additional routes explaining the too-big bandwidth utilization.