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.

CBQOS default settings are Ingress

FormerMember
FormerMember

I have 3 points to make in this posting.

As probably most customers can attest marking or re-marking your traffic on egress from your WAN Routers is the de facto standard....if you are an enterprise customer. 

Point 1

Your tool has QOS setup as everything being a default Ingress setting.  What this effectively means is the the CBQOS drill downs results in no data to be seen and the requirement to change the selection to BOTH or Egress each and every time for every chart element or drill down.  It has no memory of your last selection nor are there any selectable elements.

Point 2

On this same topic, the default Component viewable on the main NTA page has these settings:

 

You will notice that for Traffic In and Traffic Out there is no data.  Yet when you drill through on an individual component..you will indeed see the traffic.

 \

I suspect this is some sort of by design decision.

Point 3

Lastly the only way I was able to include anything at all about CBQOS on any page at all was by using an IFRAME..there is no selectable stuff nor Reports written etc.  For me the very fact the CBQOS lives in Netflow is weird.  It's a standard MIB that is SNMP pollable and should be available as part of the Universal device poller.  Or IPSLA/VOIP module or something other than NTA.

  • Hello andrewbrill,

    to answer you questions:

    ad 1. You're right we have some default direction settings for views which can't be easily changed. We'll consider this as feature of future version of NTA.

    ad 2. Yes, this is by design. But if you expand these nodes on sources you will see list of interfaces and their traffic in/out, same as on NetFlow Node Details. This is because we show information about current traffic just for interfaces.

    ad 3. You should be able to include CBQoS resources on NTA or NPM interface details view (by customize page). You can't put them on any other view, since we require interface as source. On which page do you want to put it? As for reports, we currently don't provide any, can you give an example what reports would be useful for you? And do you miss some other CBQoS-related functionality?

    Thanks for feedback, we appreciate that!

  • FormerMember
    0 FormerMember in reply to Martin.Krivanek

    In response to question 3:  Most useful report would be a Top QOS Queue Discards report and TOPx Interface by Queue type.

  • Thanks, could you please explain bit more what would these reports contain? First one should show top interfaces by bytes dropped? What about the second, what do you mean by queue type, traffic policy?

    Regards,

    Martin

  • FormerMember
    0 FormerMember in reply to Martin.Krivanek

    Traditionally, the areas of keen interest for QOS reporting are related to Queue Drops per Queue.  As an example often most people have either a 4 class or 6 Class queuing using Class Based Weighted Fair Queueing.  There can be drops on any of these queues.  This is usually measured by a count of packets.   Often these queues are also called traffic policies. 

  • Thanks, I'll forward this information to PMs so we can consider this in future NTA release.

    Regards,

    Martin

  • I could not agree more with the request for some reports on CBQoS.  Case example.  I am right now (which is why im searching THWACK) researching a class of service used for some site to site replication between 2 of my data centers.  The replication is now falling behind, which suggests to me we need to allocate more bandwidth across our oc12's.  If I could see CBQoS drops over a period of time by class of service I would know right away when the drops (which we know we have, some are actually desired) increase.  We could then correlate that to what file share is replicating at that time and how large the backup is.  Tada, QoS could be optimized.  As it is now, I’m digging for clues.  Still a huge fan, but we need some CBQoS historical reports!   

  • Great feedback guys!

    I've got these items logged and we will consider them for a future release.

  • I added the CBQOS resources to the Netflow Interface view. Doing that allows you to change the time window and look back at an event that happen as much as a few days back in time. VERY helpful! I wish I would have thought of this before!

     

              Dan Goodale - Triwest Healthcare