[c-nsp] random-detect drop counts
Tony
td_miles at yahoo.com
Sun Nov 6 19:27:46 EST 2011
Hi all,
I'm testing with the "random-detect" command to use WRED for congestion avoidance.
I've enabled it under "class-default" in a child policy, where the parent policy is a shaper to shape traffic to the purchased speed of the tail link.
My question is in regard to the counting of drops by the policy. Looking at the output below the total drop count for class-default is 4009. If you then look at the drop counts for both random drop (1622) and tail drop (219) the sum of these does NOT equal the total drop count (it's less than half).
So if a packet is not getting dropped by either random or tail drop, then how/where is it getting dropped ?
There are other classes in the policy-map with some reserved bandwidth and they aren't showing any drops (but I didn't enable WRED on them either).
If it matters HW/SW is 7609 with SIP-400 & GigE-SPA running 12.2(33)SRD4.
Class-map: class-default (match-any)
1125330 packets, 1223563563 bytes
30 second offered rate 2833000 bps, drop rate 2000 bps
Match: any
queue limit 360 packets
(queue depth/total drops/no-buffer drops) 0/4009/0
(pkts output/bytes output) 1122027/1219085597
Exp-weight-constant: 9 (1/512)
Mean queue depth: 8 packets
class Transmitted Random drop Tail drop Minimum Maximum Mark
pkts/bytes pkts/bytes pkts/bytes thresh thresh prob
0 1122025/1219085373 1622/2049685 219/286707 90 180 1/10
1 0/0 0/0 0/0 101 180 1/10
2 0/0 0/0 0/0 112 180 1/10
3 0/0 0/0 0/0 123 180 1/10
4 0/0 0/0 0/0 135 180 1/10
5 0/0 0/0 0/0 146 180 1/10
6 0/0 0/0 0/0 157 180 1/10
7 0/0 0/0 0/0 168 180 1/10
Thanks,
Tony.
More information about the cisco-nsp
mailing list