Data Plane Policing

This chapter contains the following sections:

Overview of Data Plane Policing

Use data plane policing (DPP) to manage bandwidth consumption on Cisco Application Centric Infrastructure (ACI) fabric access interfaces. DPP policies can apply to egress traffic, ingress traffic, or both. DPP monitors the data rates for a particular interface. When the data rate exceeds user-configured values, marking or dropping of packets occurs immediately. Policing does not buffer the traffic; therefore, the transmission delay is not affected. When traffic exceeds the data rate, the Cisco ACI fabric can either drop the packets or mark QoS fields in them.

Before the 3.2 release, the standard behavior for the policer was to be per-EPG member in the case of DPP policy being applied to the EPG, while the same policer was allocated on the leaf switch for the Layer 2 and Layer 3 case. This distinction was done because the DPP policer for Layer 2/Layer 3 case was assumed to be per-interface already, hence it was assumed different interfaces might get different ones. While the per-EPG DPP policy was introduced, it was clear that on a given leaf switch, several members could be present and therefore the policer it made sense to be per-member in order to avoid unwanted drops.

Starting with release 3.2, a clear semantic is given to the Data Plane Policer policy itself, as well as a new flag introducing the sharing-mode setting as presented in the CLI. Essentially, there is no longer an implicit behavior, which is different if the Data Plane Policer is applied to Layer 2/Layer 3 or to per-EPG case. Now the user has the control of the behavior. If the sharing-mode is set to shared, then all the entities on the leaf switch referring to the same Data Plane Policer, will share the same hardware policer. If the sharing-mode is set to dedicated then there would be a different HW policer allocated for each Layer 2 or Layer 3 or EPG member on the leaf switch. The policer is then dedicated to the entity that needs to be policed.

DPP policies can be single-rate, dual-rate, and color-aware. Single-rate policies monitor the committed information rate (CIR) of traffic. Dual-rate policers monitor both CIR and peak information rate (PIR) of traffic. In addition, the system monitors associated burst sizes. Three colors, or conditions, are determined by the policer for each packet depending on the data rate parameters supplied: conform (green), exceed (yellow), or violate (red).

Typically, DPP policies are applied to physical or virtual layer 2 connections for virtual or physical devices such as servers or hypervisors, and on layer 3 connections for routers. DPP policies applied to leaf switch access ports are configured in the fabric access (infra) portion of the Cisco ACI fabric, and must be configured by a fabric administrator. DPP policies applied to interfaces on border leaf switch access ports (l3extOut or l2extOut) are configured in the tenant (fvTenant) portion of the Cisco ACI fabric, and can be configured by a tenant administrator.

The data plane policer can also be applied on an EPG so that traffic that enters the Cisco ACI fabric from a group of endpoints are limited per member access interface of the EPG. This is useful to prevent monopolization of any single EPG where access links are shared by various EPGs.

Only one action can be configured for each condition. For example, a DPP policy can to conform to the data rate of 256000 bits per second, with up to 200 millisecond bursts. The system applies the conform action to traffic that falls within this rate, and it would apply the violate action to traffic that exceeds this rate. Color-aware policies assume that traffic has been previously marked with a color. This information is then used in the actions taken by this type of policer.

For information about traffic storm control, see the Cisco APIC Layer 2 Networking Configuration Guide.

Guidelines and Limitations

The following are the guidelines and limitations for configuring data plane policing:

  • The Data plane policer (DPP) does not police the packets transmitted from CPU and CPU bound packets on ACI fabric access interfaces.

  • The Dedicated Policer sharing mode is only supported for EPG level policer, not supported for physical interfaces, Layer 2 interfaces nor Layer 3 interfaces.

  • Egress traffic policing is not supported on Fabric Extender (FEX) ports

The following are the guidelines and limitations for the policer mode options: Bit policer mode (Bits-Per-Seconds: BPS) and Packet Policer mode (Packet-Per-Seconds: PPS)

  • In egress direction, both BPS and PPS policer modes are supported. (PPS policer mode requires Cisco ACI 6.1(2) or later)

  • In egress direction, PPS policer mode is supported only from FX switch onwards.

  • In ingress direction, only BPS policer mode is supported.

  • Policer statistics are available only in the policer mode that it is currently configured in. When you change the policer mode, the previously accumulated statistics are deleted.

The following are the guidelines and limitations for configuring Conform Action and Violate Action:

  • In ingress direction, the following actions are supported:

    • Conform Action: Drop and Transmit

    • Violate Action: Drop, Mark and Transmit

  • In egress direction, the following actions are supported:

    • Conform Action: Drop and Transmit

    • Violate Action: Drop

The following are guidelines and limitations for EPG policing:

  • Feature support begins with switch models ending in EX/FX (example: N9K-C93180YC-EX) and subsequent models.

  • Egress traffic policing is not supported on the EPG level policer.

  • Policer type 2R3C is not supported.

  • Policer is not supported when intra-EPG isolation is enforced in EPG.

  • Statistics and considerations for tuning include:

    • Awareness of packets that are dropped/allowed is important to know to mitigate issues or for overuse of resources.

    • Statistics are provided in the GUI using the statistics infrastructure. Statistics are exported through the REST API as for any statistic in the Cisco ACI fabric.

    • Statistics are available on per-EPG member, and are useful if the Data Plane Policer policy is of type dedicated, otherwise the statistics reflect the statistics of all the ports using it on the leaf switch.

  • In certain cases, such as when frames goes through FCoE supported devices, these get classified into the no drop FCoE class. In FCoE devices, this can cause drop off packets when the packet length is higher than the allowed 2184 bytes.

Configuring Data Plane Policing for Layer 2 Interface Using the GUI

Before you begin

The tenant, VRF, and external routed network where you configure the Data Plane Policing policy must be already created.

To apply the Layer 2 Data Plane Policing policy, the policy must be added to a policy group and the policy group must be mapped to an interface profile.

Procedure


Step 1

On the menu bar, choose Fabric > Access Policies.

In only the 3.2(1) release, the menu bar path is Fabric > External Access Policies

Step 2

In the Navigation pane, choose Policies > Interface > Data Plane Policing.

Step 3

Right-click Data Plane Policing Policing, and click Create a Data Plane Policing Policy.

Step 4

In the Create a Data Plane Policing Policy dialog box, in the Name field, enter a name for the policy.

Step 5

For Administrative State, choose enabled.

Step 6

For BGP Domain Policer Mode, choose either Bit Policer or Packet Policer.

Step 7

For Type, choose 1 Rate 2 Color or 2 Rate 3 Color.

Switch models ending in EX/FX (for example: N9K-C93180YC-EX) and subsequent models do not support 2 Rate 3 Color.

Step 8

For Conform Action, choose an action.

This choice defines an actions for traffic that conforms with certain conditions.

  • Drop: Drops the packets if the conditions are met.

  • Mark: Marks the packets if the conditions are met.

  • Transmit: Transmits the packets if the conditions are met.

Step 9

If for Conform Action you chose Mark, perform the following substeps:

  1. For Conform mark CoS, enter the class of service for packets that conformed with the conditions.

  2. For Conform mark dscp, enter the differentiated services code point (DSCP) for packets that conformed with the conditions.

Step 10

The administrator can configure the CoS and DSCP values in the Conform and Violate fields.

Step 11

If for Type you chose 2 Rate 3 Color, then for Exceed Action, choose an action.

This choice defines an actions for traffic that exceeds certain conditions.

  • Drop: Drops the packets if the conditions are met.

  • Mark: Marks the packets if the conditions are met.

  • Transmit: Transmits the packets if the conditions are met.

Step 12

If for Exceed Action you chose Mark, perform the following substeps:

  1. For Exceed mark CoS, enter the class of service for packets that exceeded the conditions.

  2. For Exceed mark dscp, enter the differentiated services code point (DSCP) for packets that exceeded the conditions.

Step 13

For Violate Action, choose an action.

This choice defines an actions for traffic that violates to certain conditions.

  • Drop: Drops the packets if the conditions are met.

  • Mark: Marks the packets if the conditions are met.

  • Transmit: Transmits the packets if the conditions are met.

Step 14

If for Violate Action you chose Mark, perform the following substeps:

  1. For Violate mark CoS, enter the class of service for packets that violated the conditions.

  2. For Violate mark dscp, enter the differentiated services code point (DSCP) for packets that violated the conditions.

Step 15

For Sharing Mode, choose Shared Policer.

Shared Policer mode allows you to apply the same policing parameters to several interfaces simultaneously. The Dedicated Policer mode is not supported for Layer 2 interfaces.

Step 16

For Rate, enter the rate at which to allow packets are allowed into the system and choose the unit per packet.

Step 17

For Burst, enter the number of packets allowed at the line rate during a burst and choose the unit per packet.

Step 18

If for Type you chose 2 Rate 3 Color, perform the following substeps:

  1. For Peak Rate, enter the peak information rate, which is the rate above which data traffic is negatively affected, and choose the unit per packet.

  2. For Excessive Burst, enter the size that a traffic burst can reach before all traffic exceeds the peak information rate, and choose the unit per packet.

Step 19

Click Submit.


Configuring Data Plane Policing for Layer 3 Interface Using the APIC GUI

Before you begin

The tenant, VRF, and external routed network where you configure the Data Plane Policing policy is already created.

The Data Plane Policing policy must be added to a policy group and the policy group mapped to an interface profile to apply the L3 DPP policy.

Procedure


Step 1

In the Navigation pane, click on Tenant_name > Networking > External Routed Network > Network_name > Logical Node Profiles > Logical Node Profile_name > Logical Interface Profiles, and perform the following actions.

  1. Right-click on Logical Interface Profiles, and select Create Interface Profile.

  2. In the Create Interface Profile dialog box, in the Name field, enter a name for the profile.

  3. Next to Ingress Data Plane Policing Policy, select Create Data Plane Policing Policy.

  4. In the Name field, enter a name for the policy.

  5. In the Administrative State field, click enabled.

  6. Next to Policer Mode, select a button for either Bit Policer or Packet Policer.

  7. Next to Type, select a button for 1 Rate 2 Color or 2 Rate 3 Color.

Switch models ending in EX/FX (for example: N9K-C93180YC-EX) and subsequent models don't support 2 Rate 3 Color).

  1. The administrator can configure the CoS and DSCP values in the Conform and Violate fields.

  2. In the Sharing Mode field, select the policer mode.

    Note

     

    Shared Policer Mode allows you to apply the same policing parameters to several interfaces simultaneously.

  3. Next to the Burst, Excessive Burst and Rate fields, select the drop down arrow to set the per packet rate for 1 Rate 2 Color policy type.

    Note

     

    For 2 Rate 3 Color policy type, the Peak Rate field is added.

  4. Click Submit.

Step 2

Expand the Routed Interfaces table, in the Path field navigate to the interface to apply the policy and perform the following actions:

  1. Next to IPv4/Ipv6 Preferred Address, enter a subnet IP address.

  2. Click OK.

  3. Click on the SVI tab and expand, in the Path field navigate to the interface to apply the policy.

  4. Next to Encap, enter the VLAN name.

  5. Next to IPv4/Ipv6 Preferred Address, enter a subnet IP address.

  6. Click OK.

  7. Expand the Routed Sub-Interfaces tab, and follow the same configuration steps as for the Routed Interfaces.

  8. Click OK. This completes DPP configuration for L3.


Configuring Data Plane Policing Using the REST API

To police the Layer 2 traffic coming in to the leaf switch:

<!--  api/node/mo/uni/.xml  -->
<infraInfra>
<qosDppPol name="infradpp5" burst="2000" rate="2000" be="400" sharingMode="shared"/>
<!--
 List of nodes. Contains leaf selectors. Each leaf selector contains list of node blocks
-->
<infraNodeP name="leaf1">
<infraLeafS name="leaf1" type="range">
<infraNodeBlk name="leaf1" from_="101" to_="101"/>
</infraLeafS>
<infraRsAccPortP tDn="uni/infra/accportprof-portselector1"/>
</infraNodeP>
<!--
 PortP contains port selectors. Each port selector contains list of ports. It
       also has association to port group policies 
-->
<infraAccPortP name="portselector1">
<infraHPortS name="pselc" type="range">
<infraPortBlk name="blk" fromCard="1" toCard="1" fromPort="48" toPort="49"></infraPortBlk>
<infraRsAccBaseGrp tDn="uni/infra/funcprof/accportgrp-portSet2"/>
</infraHPortS>
</infraAccPortP>
<!--  FuncP contains access bundle group policies  -->
<infraFuncP>
<infraAccPortGrp name="portSet2">
<infraRsQosIngressDppIfPol tnQosDppPolName="infradpp5"/>
</infraAccPortGrp>
</infraFuncP>
</infraInfra>

To police the Layer 2 traffic going out of the leaf switch:

<!--  api/node/mo/uni/.xml  -->
<infraInfra>
<qosDppPol name="infradpp2" burst="4000" rate="4000"/>
<!--
 List of nodes. Contains leaf selectors. Each leaf selector contains list of node blocks
-->
<infraNodeP name="leaf1">
<infraLeafS name="leaf1" type="range">
<infraNodeBlk name="leaf1" from_="101" to_="101"/>
</infraLeafS>
<infraRsAccPortP tDn="uni/infra/accportprof-portselector2"/>
</infraNodeP>
<!--
 PortP contains port selectors. Each port selector contains list of ports. It
       also has association to port group policies 
-->
<infraAccPortP name="portselector2">
<infraHPortS name="pselc" type="range">
<infraPortBlk name="blk" fromCard="1" toCard="1" fromPort="37" toPort="38"></infraPortBlk>
<infraRsAccBaseGrp tDn="uni/infra/funcprof/accportgrp-portSet2"/>
</infraHPortS>
</infraAccPortP>
<!--  FuncP contains access bundle group policies  -->
<infraFuncP>
<infraAccPortGrp name="portSet2">
<infraRsQosEgressDppIfPol tnQosDppPolName="infradpp2"/>
</infraAccPortGrp>
</infraFuncP>
</infraInfra>

To police the Layer 3 traffic coming in to the leaf switch:

<!--  api/node/mo/uni/.xml  -->
<fvTenant name="dppTenant">
<qosDppPol name="gmeo" burst="2000" rate="2000"/>
<l3extOut name="Outside">
<l3extInstP name="extroute"/>
<l3extLNodeP name="borderLeaf">
<l3extRsNodeL3OutAtt tDn="topology/pod-1/node-101" rtrId="10.0.0.1">
<ipRouteP ip="0.0.0.0">
<ipNexthopP nhAddr="192.168.62.2"/>
</ipRouteP>
</l3extRsNodeL3OutAtt>
<l3extLIfP name="portProfile">
<l3extRsPathL3OutAtt addr="192.168.40.1/30" ifInstT="l3-port" tDn="topology/pod-1/paths-101/pathep-[eth1/40]"/>
<l3extRsPathL3OutAtt addr="192.168.41.1/30" ifInstT="l3-port" tDn="topology/pod-1/paths-101/pathep-[eth1/41]"/>
<l3extRsIngressQosDppPol tnQosDppPolName="gmeo"/>
</l3extLIfP>
</l3extLNodeP>
</l3extOut>
</fvTenant>

To police the Layer 3 traffic going out of the leaf switch:

<!--  api/node/mo/uni/.xml  -->
<fvTenant name="dppTenant">
<qosDppPol name="gmeo" burst="2000" rate="2000"/>
<l3extOut name="Outside">
<l3extInstP name="extroute"/>
<l3extLNodeP name="borderLeaf">
<l3extRsNodeL3OutAtt tDn="topology/pod-1/node-101" rtrId="10.0.0.1">
<ipRouteP ip="0.0.0.0">
<ipNexthopP nhAddr="192.168.62.2"/>
</ipRouteP>
</l3extRsNodeL3OutAtt>
<l3extLIfP name="portProfile">
<l3extRsPathL3OutAtt addr="192.168.40.1/30" ifInstT="l3-port" tDn="topology/pod-1/paths-101/pathep-[eth1/40]"/>
<l3extRsPathL3OutAtt addr="192.168.41.1/30" ifInstT="l3-port" tDn="topology/pod-1/paths-101/pathep-[eth1/41]"/>
<l3extRsEgressQosDppPol tnQosDppPolName="gmeo"/>
</l3extLIfP>
</l3extLNodeP>
</l3extOut>
</fvTenant>

Configuring Data Plane Policing Using NX-OS Style CLI

Procedure


Step 1

Configure a Layer 2 port to carry one EPG.

Example:

apic1# conf t
apic1(config)# vlan-domain test
apic1(config-vlan)# vlan 1000-2000
apic1(config-vlan)# exit
apic1(config)# leaf 101
apic1(config-leaf)# interface ethernet 1/10
apic1(config-leaf-if)# vlan-domain member test
apic1(config-leaf-if)# exit
apic1(config-leaf)# exit
apic1(config)# tenant test1
apic1(config-tenant)# vrf context v1
apic1(config-tenant-vrf)# exit
apic1(config-tenant)# bridge-domain bd1
apic1(config-tenant-bd)# vrf member v1
apic1(config-tenant-bd)# exit
apic1(config-tenant)# application ap1
apic1(config-tenant-app)# epg e1
apic1(config-tenant-app-epg)# bridge-domain member bd1
apic1(config-tenant-app-epg)# exit
apic1(config-tenant-app)# exit
apic1(config-tenant)# exit
apic1(config)# leaf 101
apic1(config-leaf)# interface ethernet 1/10
apic1(config-leaf-if)# switchport trunk allowed vlan 1001 tenant test1 application ap1 epg e1
apic1(config-leaf-if)# switchport trunk allowed vlan 1501 tenant test1 application ap1 epg e1
# Now the port leaf 101 ethernet 1/10 carries two vlan mapped both to the same Tenant/Application/EPG
apic1(config-leaf-if)# exit
apic1(config-leaf)# exit
  1. Create a policy-map to apply to the interface.

    Example:

    apic1(config)# policy-map type data-plane qosTest
    apic1(config-pmap-dpp)# set burst 2400 mega
    apic1(config-pmap-dpp)# set cir 70 mega

    apic1(config-pmap-dpp)# set sharing-mode shared
    apic1(config-pmap-dpp)# exit
    apic1(config)# leaf 101
    apic1(config-leaf)# interface ethernet 1/10
    apic1(config-leaf-if)# service-policy type data-plane input qosTest
    apic1(config-leaf-if)# exit
    apic1(config-leaf)# exit
    apic1(config)# policy-map type data-plane qosTest2
    apic1(config-pmap-dpp)# set cir 78 mega
    apic1(config-pmap-dpp)# exit
    apic1(config)# leaf 101
    apic1(config-leaf)# interface ethernet 1/10
    apic1(config-leaf-if)# service-policy type data-plane output qosTest2
    apic1(config-leaf-if)# end
  2. Visualize the policy configured.

    Example:

    apic1# show policy-map type data-plane infra
    Type data-plane policy-maps
    ====================
    Global Policy
    policy-map type data-plane default
        set burst unspecified
        set conform-cos-transmit unspecified
        set conform-dscp-transmit unspecified
        set conform transmit
        set excessive-burst unspecified
        set exceed-cos-transmit unspecified
        set exceed-dscp-transmit unspecified
        set exceed drop
        set mode byte
        set pir 0
        set cir 78 mega
        set type 1R2C
        set violate-cos-transmit unspecified
        set violate-dscp-transmit unspecified
        set violate drop
    Global Policy
    policy-map type data-plane qosTest
        set burst 2400 mega
        set cir 78 mega
        set conform-cos-transmit unspecified
        set conform-dscp-transmit unspecified
        set conform transmit
        set excessive-burst unspecified
        set exceed-cos-transmit unspecified
        set exceed-dscp-transmit unspecified
        set exceed drop
        set mode byte
        set pir 0
        set type 1R2C
        set violate-cos-transmit unspecified
        set violate-dscp-transmit unspecified
        set violate drop
    Global Policy
    policy-map type data-plane qosTest2
        set burst unspecified
        set conform-cos-transmit unspecified
        set conform-dscp-transmit unspecified
        set conform transmit
        set excessive-burst unspecified
        set exceed-cos-transmit unspecified
        set exceed-dscp-transmit unspecified
        set exceed drop
        set mode byte
        set pir 0
        set cir 78 mega
        set type 1R2C
        set violate-cos-transmit unspecified
        set violate-dscp-transmit unspecified
        set violate drop
  3. Show running-config.

    Example:

    apic1# show runn policy-map
    # Command: show running-config policy-map
    # Time: Fri Jan 29 19:26:18 2016
      policy-map type data-plane default
        exit
      policy-map type data-plane qosTest
        set burst 2400 mega
        set cir 78 mega
        no shutdown
        exit
      policy-map type data-plane qosTest2
        set cir 78 mega
        no shutdown
        exit
    apic1# show runn leaf 101
    # Command: show running-config leaf 101
    # Time: Fri Jan 29 19:26:29 2016
      leaf 101 
        interface ethernet 1/10
          vlan-domain member test
          switchport trunk allowed vlan 1501 tenant test1 application ap1 epg e1 
          service-policy type data-plane input qosTest
          service-policy type data-plane output qosTest2
          exit
        exit

Step 2

Preparation to configure Layer 3 ports.

Example:

apic1# conf t
apic1(config)# vlan-domain l3ports
apic1(config-vlan)# vlan 3000-3001
apic1(config-vlan)# exit
apic1(config)# tenant l3test1
apic1(config-tenant)# vrf context v1
apic1(config-tenant-vrf)# exit
apic1(config-tenant)# exit
apic1(config)# leaf 102
apic1(config-leaf)# vrf context tenant l3test1 vrf v1
apic1(config-leaf-vrf)# exit
# Configure a physical Layer 3 port
apic1(config-leaf)# interface ethernet 1/20
apic1(config-leaf-if)# no switchport
apic1(config-leaf-if)# vlan-domain member l3ports
apic1(config-leaf-if)# vrf member tenant l3test1 vrf v1
apic1(config-leaf-if)# ip address 56.1.1.1/24
apic1(config-leaf-if)# ipv6 address 2000::1/64  preferred
apic1(config-leaf-if)# exit
# Configure base interface for L3 subinterfaces
apic1(config-leaf)# interface ethernet 1/21
apic1(config-leaf-if)# vlan-domain member l3ports
apic1(config-leaf-if)# no switchport
apic1(config-leaf-if)# exit
# Configure a Layer 3 subinterface
apic1(config-leaf)# interface ethernet 1/21.3001
apic1(config-leaf-if)# vrf member tenant l3test1 vrf v1
apic1(config-leaf-if)# ip address 60.1.1.1/24
apic1(config-leaf-if)# ipv6 address 2001::1/64 preferred
apic1(config-leaf-if)# exit
# Configure a Switched Vlan Interface
apic1(config-leaf)# interface vlan 3000
apic1(config-leaf-if)# vrf member tenant l3test1 vrf v1
apic1(config-leaf-if)# ip address 70.1.1.1/24
apic1(config-leaf-if)# ipv6 address 3000::1/64 preferred
apic1(config-leaf-if)# exit
apic1(config-leaf)# exit
  1. Configure the policer in the tenant for Layer 3 usage.

    Example:

    apic1(config)# tenant l3test1
    apic1(config-tenant)# policy-map type data-plane iPol
    apic1(config-tenant-pmap-dpp)# set cir 56 mega
    apic1(config-tenant-pmap-dpp)# set burst 2000 kilo
    apic1(config-tenant-pmap-dpp)# exit
    apic1(config-tenant)# policy-map type data-plane ePol
    apic1(config-tenant-pmap-dpp)# set burst 2000 kilo
    apic1(config-tenant-pmap-dpp)# set cir 56 mega
    apic1(config-tenant-pmap-dpp)# exit
    apic1(config-tenant)# exit
  2. Apply policer on a Layer 3 interface

    Example:

    apic1(config)# leaf 102
    apic1(config-leaf)# interface ethernet 1/20
    apic1(config-leaf-if)# service-policy type data-plane input iPol
    apic1(config-leaf-if)# service-policy type data-plane output ePol
    apic1(config-leaf-if)# exit
    apic1(config-leaf)# interface ethernet 1/21.3001
    apic1(config-leaf-if)# service-policy type data-plane input iPol
    apic1(config-leaf-if)# service-policy type data-plane output ePol
    apic1(config-leaf-if)# exit
    apic1(config-leaf)# interface vlan 3000
    apic1(config-leaf-if)# service-policy type data-plane input iPol
    apic1(config-leaf-if)# service-policy type data-plane output ePol
    apic1(config-leaf-if)# end
  3. Show commands for policers used on a Layer 3 interface.

    Example:

    apic1# show tenant l3test1 policy-map type data-plane
    Type data-plane policy-maps
    ====================
    Policy in Tenant: l3test1
    policy-map type data-plane ePol
        set burst 2000 kilo
        set conform-cos-transmit unspecified
        set conform-dscp-transmit unspecified
        set conform transmit
        set excessive-burst unspecified
        set exceed-cos-transmit unspecified
        set exceed-dscp-transmit unspecified
        set exceed drop
        set mode byte
        set pir 0
        set cir 56 mega
        set type 1R2C
        set violate-cos-transmit unspecified
        set violate-dscp-transmit unspecified
        set violate drop
    Policy in Tenant: l3test1
    policy-map type data-plane iPol
        set burst 2000 kilo
        set burst unspecified
        set conform-cos-transmit unspecified
        set conform-dscp-transmit unspecified
        set conform transmit
        set excessive-burst unspecified
        set exceed-cos-transmit unspecified
        set exceed-dscp-transmit unspecified
        set exceed drop
        set mode byte
        set pir 0
        set cir 56 mega
        set type 1R2C
        set violate-cos-transmit unspecified
        set violate-dscp-transmit unspecified
        set violate drop
  4. Show running-config for policers used for Layer 3.

    Example:

    apic1# show runn tenant l3test1
    # Command: show running-config tenant l3test1
    # Time: Fri Jan 29 19:48:20 2016
      tenant l3test1
        vrf context v1
          exit
        policy-map type data-plane ePol
          set burst 2000 kilo
          set cir 56 mega
          no shutdown
          exit
        policy-map type data-plane iPol
          set burst 2000 kilo
          set cir 56 mega
          no shutdown
          exit
        exit
    apic1# show running-config leaf 102
    # Command: show running-config leaf 102
    # Time: Fri Jan 29 19:48:33 2016
      leaf 102
        vrf context tenant l3test1 vrf v1
          exit
        interface vlan 3000
          vrf member tenant l3test1 vrf v1
          ip address 70.1.1.1/24
          ipv6 address 3000::1/64 preferred
          bfd ip tenant mode
          bfd ipv6 tenant mode
          service-policy type data-plane input iPol
          service-policy type data-plane output ePol
          exit
        interface ethernet 1/20
          vlan-domain member l3ports
          no switchport
          vrf member tenant l3test1 vrf v1
          ip address 56.1.1.1/24
          ipv6 address 2000::1/64 preferred
          bfd ip tenant mode
          bfd ipv6 tenant mode
          service-policy type data-plane input iPol
          service-policy type data-plane output ePol
          exit
        interface ethernet 1/21
          vlan-domain member l3ports
          no switchport
          bfd ip tenant mode
          bfd ipv6 tenant mode
          exit
        interface ethernet 1/21.3001
          vrf member tenant l3test1 vrf v1
          ip address 60.1.1.1/24
          ipv6 address 2001::1/64 preferred
          bfd ip tenant mode
          bfd ipv6 tenant mode
          service-policy type data-plane input iPol
          service-policy type data-plane output ePol
        exit
        exit
    apic1#

Data Plane Policing at the Endpoint Group Level

Data Plane Policing (DPP) can be applied to an endpoint group (EPG). The policing of the traffic is applied to all the EPG members on every leaf switch where the EPG is deployed.

Prior to the 3.2(1) release, each EPG member had its own policer. Beginning in the 3.2(1) release, the behavior is dependent on the sharing-mode property (if configured through the CLI or GUI) on the Data Plane Policer. If that is set to dedicated, then the situation is similar to before the 3.2(1) release. If the sharing-mode is set to shared, then all the members in the same slice using the same Data Plane Policer policy use the hardware policer on the leaf switch.

For example, an EPG has the following members:

  • Leaf 101, Eth1/1, vlan-300

  • Leaf 101, Eth1/2, vlan-301

  • Leaf 102, Eth1/2, vlan-500

In this case, each member will limit the traffic according to the policer, independent from the other members. If the Data Plane Policer has the sharing-mode set to shared, then all the members in the same slice above use only one policer on the leaf switch.

The Data Plane Policer works independently on Leaf 101 and Leaf 102 if the sharing-mode is set to dedicated. For example:

  • Policer-A (100Mbps policing) is applied to EPG1 (Leaf101 e1/1 vlan-300 and e1/2 vlan-301. Leaf 102 e1/2 vlan-500)

  • Leaf 101: police traffic at the EPG1 level, which is applied to traffic through E1/1 vlan-300 and E1/2 vlan-301 (100Mbps for each interface).

  • Leaf 102: police traffic at the EPG1 level, which is applied to traffic through E1/2 vlan-500 (another 100Mbps for each interface).

The total is up to 300Mbps for EPG1.

If the sharing-mode is set to shared, 100Mbps is shared across EPGs using the same policer if the interfaces are in the same slice. For example:

  • Policer-A (100Mbps policing) applied to EPG1 and EPG2.

  • Leaf 101: police traffic at EPG1 and EPG2 in total.

  • Leaf 102: police traffic at EPG1 and EPG2 in total.

The total is up to 200Mbps for EPG1 and EPG2 if the interfaces are in the same slice.

The following are limitations for Data Plane Policing at the EPG level:

  • EPG policer feature is supported with switch models that have -EX, -FX, or later suffixes in the product ID.

  • Egress traffic policing is not supported for the EPG level policer.

  • Policer mode Packet-per-second is not supported.

  • Policer type 2R3C is not supported in EPG policer.

  • Policer is not supported when intra-EPG isolation-enforced is applied to the EPG.

  • The scale limit allows for 128 EPG policers supported per node.

Configuring Data Plane Policing at the Endpoint Group Level Using CLI

SUMMARY STEPS

  1. Define the policer:

DETAILED STEPS


Define the policer:

Example:

apic1# conf t
apic1(config)# vlan-domain test
apic1(config-vlan)# vlan 1000-2000
apic1(config-vlan)# exit
apic1(config)# leaf 101
apic1(config-leaf)# interface ethernet 1/10
apic1(config-leaf-if)# vlan-domain member test
apic1(config-leaf-if)# exit
apic1(config-leaf)# exit
apic1(config)# tenant test1
apic1(config-tenant)# vrf context v1
apic1(config-tenant-vrf)# exit
apic1(config-tenant)# bridge-domain bd1
apic1(config-tenant-bd)# vrf member v1
apic1(config-tenant-bd)# exit
apic1(config)# policy-map type data-plane pol1
apic1(config-pmap-dpp)# set burst 2400 mega
apic1(config-pmap-dpp)# set cir 78 mega
apic1(config-pmap-dpp)# exit
apic1(config-tenant)# application ap1
apic1(config-tenant-app)# epg e1
apic1(config-tenant-app-epg)# bridge-domain member db1
apic1(config-tenant-app-epg)# service-policy type data-plane poll
apic1(config-tenant-app-epg)# exit
apic1(config-tenant-app)# exit
apic1(config-tenant)# exit
apic1(config)# leaf 101
apic1(config-leaf)# interface ethernet 1/10
apic1(config-leaf-if)# switchport trunk allowed vlan 1001 tenant test1 application ap1 epg e1
apic1(config-leaf-if)# exit
apic1(config-leaf)# exit

Configuring Data Plane Policing at the Endpoint Group Level Using the APIC GUI

Procedure


In the Tenants pane, click on Tenant_name > Policies > Protocol > Data Plane Policing. Right-click on Data Plane Policing to Create Data Plane Policing Policy.

  1. In the Name field, enter a name for the policy.

  2. In the Administrative State field, click enabled.

  3. Next to Policer Mode, select a button for either Bit Policer or Packet Policer.

  4. Next to Type, select a button for 1 Rate 2 Color.

  5. For Conform Action, select Drop, Mark, or Transmit.

  6. The administrator can configure the CoS and DSCP values in the Conform and Violate fields.

  7. Next to the Burst, Excessive Burst and Rate fields, click the drop down arrow to select from the following:

    • Bytes/Packets

    • Kilo Bytes/Packets

    • Mega Bytes/Packets

    • Giga Bytes/Packets

    • Milli Seconds

    • Micro Seconds


Configuring Data Plane Policing at the Endpoint Group Level Using Rest API

To police the traffic coming into the leaf switch:

<!-- api/node/mo/.xml -->
<polUni>
  <fvTenant  name="t1">                                                                                                                                               
    <qosDppPol name="gmeo" burst="2000" rate="2000"/>
    <fvAp name="ap1">
      <fvAEPg name="ep1">
        <fvRsDppPol tnQosDppPolName="gmeo"/>
      </fvAEPg>
    </fvAp> 
  </fvTenant>
</polUni>

Accessing Statistics for the Data Plane Policer at the Endpoint Group Level in the GUI

DPP at the EPG level is used to police traffic at the EPG member level. As such, statistics are integral in ensuring the policer is dropping substantial traffic. Statistics are reported at the EPG member level for fine granularity.

Procedure


Step 1

In the Tenants pane, click on Tenant_name > Application EPGs > EPG Members > Static EPG Members .

Step 2

Select a node.

Step 3

Click Select Stats.

  1. Select a Sampling Interval unit of time.

  2. From the Available policer attributes, use the arrows to choose the attributes. You can select up to two attributes.

  3. Click Submit.


What to do next

You will see a graphical representation of the DPP statistics.