random-detect dscp-based thresholds explanation

[markschnabel]

Can someone explain why the default min thresholds for dscp are larger for the higher probability AF class than the lower probability classes?

for example:

http://www.cisco.com/en/US/docs/ios/qos/command/reference/qos_q1.html#wp1052917

dscp 10 has min of 24

dscp 12 has min of 28

thanks in advance

[Sergei Vasilenko]

Hello,

Indeed, looks like this is a mistake in this table.

Should be another way around - af11(dscp10) - min thr 32, af12(dscp12) - min thr 28, af13(dscp14) - min thr 24.

Real values could be checked via "show policy-map int" on CLI, like shown at the following example link:

DSCP Value Configuration Verification: Example

http://www.cisco.com/en/US/docs/ios/qos/configuration/guide/config_wred_ps6441_TSD_Products_Configuration_Guide_Chapter.html#wp1002513

Router# show policy interface Serial6/3 

 Serial6/3 

  Service-policy output: test

    Class-map: c1 (match-any)

      0 packets, 0 bytes

      5 minute offered rate 0 bps, drop rate 0 bps

      Match: protocol ip

        0 packets, 0 bytes

        5 minute rate 0 bps

      Weighted Fair Queueing

        Output Queue: Conversation 265 

        Bandwidth 20 (%)

        Bandwidth 308 (kbps)

        (pkts matched/bytes matched) 0/0

        (depth/total drops/no-buffer drops) 0/0/0

         exponential weight: 9

         mean queue depth: 0

dscp     Transmitted       Random drop      Tail drop    Minimum Maximum  Mark

          pkts/bytes        pkts/bytes       pkts/bytes    thresh  thresh  prob

af11          0/0               0/0              0/0           32      40  1/10

af12          0/0               0/0              0/0           28      40  1/10

af13          0/0               0/0              0/0           24      40  1/10

af21          0/0               0/0              0/0           32      40  1/10

af22          0/0               0/0              0/0           28      40  1/10

af23          0/0               0/0              0/0           24      40  1/10

af31          0/0               0/0              0/0           32      40  1/10
...

Thanks,
Sergey