Re: The doubt about "qos-preclassify" command

From: Petr Lapukhov (petr@internetworkexpert.com)
Date: Sun Sep 03 2006 - 11:57:26 ART


Here is a sample configuration. I hope formatting won't die out :)

-------------------------

R4 and R5 are connected via FR cloud with DLCIs 405/504.

R4 is connected via subinterface Serial 0/0.1245.

Tunnel0 is used to tunnel traffic between VLAN4 to VLAN5 over
FR (VLAN4: 10.4.4.0/24 <-> VLAN5: 10.5.5.0/24).

The goal is to provide priority treatment of ICMP traffic between
VLAN4 and VLAN5. MQC FRTS is used on subinterface.

R4:

interface Tunnel0
 tunnel source Loopback0
 tunnel destination 150.1.5.5
 ip address 45.45.45.4 255.255.255.0
!

ip access-list extended VLAN4_TO_VLAN5
 permit icmp 10.4.4.0 0.0.0.255 10.5.5.0 0.0.0.255
!
class-map VPN_ICMP
 match access-group name VLAN4_TO_VLAN5
!
policy-map VPN_QOS
 class VPN_ICMP
  priority 16
!
policy-map POLICY_TO_R5
 class class-default
  shape average 64000
  service-policy VPN_QOS
!
map-class frame-relay PVC_405
 service-policy output POLICY_TO_R5
!
interface Serial 0/0.1245
 frame-relay class PVC_405
!
interface Tunnel0
 qos pre-classify

Verification:---------------------------------

Rack1R4#ping 10.5.5.5 source 10.4.4.4 repeat 100

Type escape sequence to abort.

Sending 100, 100-byte ICMP Echos to 10.5.5.5, timeout is 2 seconds:

Packet sent with a source address of 10.4.4.4

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Success rate is 100 percent (100/100), round-trip min/avg/max = 72/72/92 ms

Rack1R4#show policy-map interface serial 0/0.1245

 Serial0/0.1245: DLCI 405 -

  Service-policy output: POLICY_TO_R5

    Class-map: class-default (match-any)

      102 packets, 12964 bytes

      5 minute offered rate 2000 bps, drop rate 0 bps

      Match: any

      Traffic Shaping

           Target/Average Byte Sustain Excess Interval Increment

             Rate Limit bits/int bits/int (ms) (bytes)

            64000/64000 2000 8000 8000 125 1000

        Adapt Queue Packets Bytes Packets Bytes Shaping

        Active Depth Delayed Delayed Active

        - 0 102 12964 0 0 no

      Service-policy : VPN_QOS

        Class-map: ICMP (match-all)

          100 packets, 12800 bytes

          5 minute offered rate 2000 bps, drop rate 0 bps

          Match: access-group name VLAN4_TO_VLAN5

          Queueing

            Strict Priority

            Output Queue: Conversation 24

            Bandwidth 16 (kbps) Burst 400 (Bytes)

            (pkts matched/bytes matched) 0/0

            (total drops/bytes drops) 0/0

        Class-map: class-default (match-any)

          2 packets, 164 bytes

          5 minute offered rate 0 bps, drop rate 0 bps

          Match: any

Rack1R4#show interfaces tunnel 0

Tunnel0 is up, line protocol is up

  Hardware is Tunnel

  Internet address is 45.45.45.4/24

  MTU 1514 bytes, BW 9 Kbit, DLY 500000 usec,

     reliability 255/255, txload 28/255, rxload 28/255

  Encapsulation TUNNEL, loopback not set

  Keepalive not set

  Tunnel source 150.1.4.4 (Loopback0), destination 150.1.5.5

  Tunnel protocol/transport GRE/IP

    Key disabled, sequencing disabled

    Checksumming of packets disabled

  Tunnel TTL 255

  Fast tunneling enabled

  Tunnel transmit bandwidth 8000 (kbps)

  Tunnel receive bandwidth 8000 (kbps)

  Last input 00:00:08, output 00:00:08, output hang never

  Last clearing of "show interface" counters never

  Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 3

  Queueing strategy: fifo (QOS pre-classification)

  Output queue: 0/0 (size/max)

  5 minute input rate 1000 bits/sec, 1 packets/sec

  5 minute output rate 1000 bits/sec, 1 packets/sec

     614 packets input, 75464 bytes, 0 no buffer

     Received 0 broadcasts, 0 runts, 0 giants, 0 throttles

     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored, 0 abort

     615 packets output, 75540 bytes, 0 underruns

     0 output errors, 0 collisions, 0 interface resets

     0 output buffer failures, 0 output buffers swapped out

Petr Lapukhov, CCIE #16379
petr@internetworkexpert.com

Internetwork Expert, Inc.
http://www.InternetworkExpert.com
Toll Free: 877-224-8987
Outside US: 775-826-4344

----- Original Message -----
From: "Niche" <jackyliu419@gmail.com>
To: "Petr Lapukhov" <petr@internetworkexpert.com>
Cc: <ccielab@groupstudy.com>
Sent: Sunday, September 03, 2006 6:36 PM
Subject: Re: The doubt about "qos-preclassify" command

> Hi Petr,
>
> So, you mean I should get hit count increase for the matching ACL,
> right? If that's the case, would you mind sharing a sample
> configuration so I can count check mine.
>
> Thanks alot!
>
> Best Regards,
> Jacky
>
> On 9/3/06, Petr Lapukhov <petr@internetworkexpert.com> wrote:
>> I tried QoS pre-classify in my labs (GRE Tunnel, IPsec,
Virtual-Templates),
>> and it worked seamlessly with physical interfaces as well as with logical
>> subinterfaces.
>>
>> The only problem I once had on 12.3(14)T 3640 is that QoS Pre-Classify
>> didn't work with CEF enabled ;)
>>
>> HTH
>>
>> 2006/9/3, Niche <jackyliu419@gmail.com>:
>> >
>> Hi guys,
>>
>> The command should be "qos pre-classify", my mistake =P
>>
>> Cheers~
>> Jacky
>>
>> On 9/3/06, Niche <jackyliu419@gmail.com> wrote:
>> > Hi guys,
>> >
>> > Two sites are being connected via GRE tunnel, and using OSPF for
>> > routing information exchange.
>> >
>> > Requirement - QoS for traffic control but many of you may already know
>> > there are not much QoS can do to the GRE tunnel. So I used
>> > "qos-preclassify" command on the tunnel interface and configure
>> > appropriate QoS configuration in the service policy for the physical
>> > interfaces at both sites. The matching method is simple access-group
>> > with protocol type and ip addresses, and here come the odd...
>> >
>> > When I check the hit count of the ACLs that are being used by the
>> > service policy, the number is zero. I am sure there are matching
>> > traffic types passing throught the interface, so it make me wonder
>> > what's wrong.
>> >
>> > Ok, according to Cisco web information, only term "physical interface"
>> > is mentioned in the article. So, how about logical interface? e.g.
>> > multilink interface, FR sub-interface..etc. I confess that the service
>> > policies for both of sites are being bound on logical interfaces
>> > instead of physical one.
>> >
>> > Cheers~
>> > Jacky
>>
>> _______________________________________________________________________
>> Subscription information may be found at:
>> http://www.groupstudy.com/list/CCIELab.html
>>
>>
>>
>> --
>> Petr Lapukhov, CCIE #16379
>> petr@internetworkexpert.com
>>
>> Internetwork Expert, Inc.
>> http://www.InternetworkExpert.com
>> Toll Free: 877-224-8987
>> Outside US: 775-826-4344



This archive was generated by hypermail 2.1.4 : Sun Oct 01 2006 - 16:55:39 ART