Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Accounting of the Policer Overhead Attribute at the Interface Level

A bandwidth profile (BWP) enforces limits on bandwidth utilization according to the service level specification (SLS) that has been agreed upon by the subscriber and the service provider as part of the service level agreement (SLA).There are two types of bandwidth profiles:

  • Ingress Bandwidth Profile

  • Egress Bandwidth Profile

Need for Policer Overhead adjustment

The Metro Ethernet Forum (MEF) Carrier Ethernet (CE) 2.0 set of standards has stringent requirements for the bandwidth profile enforcement at the user network interface (UNI). MEF CE 2.0 certification compliance allows only a 2 percent deviation from UNI commited or peak rate across all frame sizes. This means that the policers should take into account the frame size at the UNI interface, including frame checksum and excluding all additional overheads that might be added inside the service provider network (such as S-VLANs). Therefore, this translates into two customer requirements:

  • Junos OS egress policers use frame length before output VLAN manipulation. If VLANs are added on output, those extra bytes will not be counted. In order to address MEF CE 2.0 requirements, adjust the length of the packet that is used for policing purposes for Junos egress policers that use frame length before output VLAN manipulation. Therefore, if VLANs are added on output, the extra bytes will not be counted.

  • In some network designs, bandwidth profile enforcement is implemented at the Layer 2 (L2) VPN Provider Edge Router and not at the Ethernet access device (EAD) terminating the physical UNI interface. The EAD typically adds an S-VLAN that identifies the port in the access network. The S-VLAN that is added is considered internal to the service provider network and typically should not be taken into account for bandwidth profile enforcement purposes at the Provider Edge device in both ingress and egress directions. This also translates into a requirement to allow adjusting the packet length used for policing purposes on ingress and egress.

In order to address these requirements, policer-overhead adjustment is defined on a per logical interface (IFL)/direction granularity, which is the range of -16 bytes to +16 bytes. The policer-overhead adjustment is applied for all the policers that take into account Layer 1 (L1) or L2 packet length that are exercised in the specified IFL/direction, including corresponding logical interface family (IFF) feature policers, and is applied only to interface or filter-based policers.

Policer Overhead Adjustment Applicability for Policers

The ingress or egress policer overhead adjustment configuration is applicable for the following types of policers on ingress or egress IFL and IFF, respectively:

  • L2 two-color and three-color policers.

  • IFL-level policers (configured at the IFL or in a filter attached to IFL).

  • Family-level policers that use L2 packet length, or policers in filters attached to L2 IFF (family bridge, vpls, ccc).

Note:

The list is applicable for all types of policers, including regular two-color policers, three-color policers, and hierarchical policers, provided that the policer operates on an L1 or L2 packet length.

Ingress policer overhead adjustment configuration is applied to any policers attached to ingress L2 routing instances.

Note:

Note that any IFF policer on the L3 family (inet inet6), which considers only L3 packet length, will not consider this adjustment. The policer overhead adjustment value (+ve or -ve) is added to the actual L2 packet length to obtain the number of bytes to charge the policer. Therefore, this is used only as an intermediate value, and does not affect actual L2 packet length. This feature is applied before VLAN normalization, and is independent of the egress-shaping-overhead or ingress-shaping-overhead configuration under class of service.