[j-nsp] QOS Mechanics - PLP and drop-profiles behaviour

[Damon Pegg]

Under test, I set my scheduler drop-profiles to be acting only against tcp
traffic.  Now, if I wind a queue up so that it is heavily congested but only
with UDP streams, I still see the sub-queue classified as 'high'
drop-priority more aggressively dealt with than the 'low' drop-priority
classified sub-queue  (I use the term sub-queue here for the diferentiation
between high and low drop priroties.)

In fact, even if I remove all drop-profiles so that allegedly everything
uses the default-drop-profile <fill-level 100 drop 100> I see the same as
above on TCP and non-tcp traffic.  Do I assume from this that the PLP status
that is used internally to differentiate the high and low traffic within a
queue (show class-of-service forwarding-table ) IS used during heavy
congestion to differentiate between the classifications..?  That is, the
high/low dichotomy isn't just for differnent WRED drop-profiles to be set?

In fact, I'm sure this is happening and although it is in fact useful to
prioritise the 'low' sub-queue during heavy congestion for all traffic,  I'd
like to know how it works so I can possibly set a non-tcp drop-profile set
to compliment the internal handling based on PLP...

Damon Pegg
Network Development
Easynet plc
t:    0207 900 7075
f:    0207 900 4443
m:    07931 406206
e:    damon.pegg at uk.easynet.net
w:    www.easynet.com