Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Enhanced Queuing DPC CoS Properties

On a Juniper Networks MX Series 5G Universal Routing Platform with Enhanced Queuing Dense Port Concentrators (DPCs), you can configure schedulers and queues. You can configure 15 VLAN sets per Gigabit Ethernet (1G) port and 255 VLAN sets per 10-Gigabit Ethernet (10G) port. The Enhanced Queuing DPC performs priority propagation from one hierarchy level to another and drop statistics are available on the Enhanced Queuing DPC per color per queue instead or just per queue.

Note:

The Enhanced Queuing DPC (EQ DPC) does not support BA classification for packets received from a Layer 3 routing interface or a virtual routing and forwarding (VRF) interface and routed to an integrated routing and bridging interface (IRB) to reach the remote end of a pseudowire connection. The EQ DPC also does not support BA classification for Layer 2 frames received from a Virtual Private LAN Service (VPLS) pseudowire connection from a remote site and routed to a Layer 3 routing interface through an IRB interface.

Juniper Networks MX Series 5G Universal Routing Platforms with Enhanced Queuing DPCs have Packet Forwarding Engines that can support up to 515 MB of frame memory, and packets are stored in 512-byte frames. Table 1 compares the major properties of the Intelligent Queuing 2 (IQ2) PIC and the Packet Forwarding Engine within the Enhanced Queuing DPC.

Table 1: IQ2 PIC and Enhanced Queuing DPC Compared

Feature

IQ2 PIC

Packet Forwarding Engine Within Enhanced Queuing DPC

Number of usable queues

8,000

16,000

Number of shaped logical interfaces

1,000 with 8 queues each.

2,000 with 8 queues each, or 4,000 with 4 queues each.

Number of hardware priorities

2

4

Priority propagation

No

Yes

Dynamic mapping

No: schedulers/port are fixed.

Yes: schedulers/port are not fixed.

Drop statistics

Per queues

Per queue per color (PLP high, low)

In addition, the Enhanced Queuing DPC features support for hierarchical weighted random early detection (WRED) and enhanced queuing on aggregated Ethernet interfaces with link protection as well.

The Enhanced Queuing DPC supports the following hierarchical scheduler characteristics:

  • Shaping at the physical interface level

  • Shaping and scheduling at the service VLAN interface set level

  • Shaping and scheduling at the customer VLAN logical interface level

  • Scheduling at the queue level

VLAN (Level 3) shaping on a 10-Gigabit Ethernet MX Series Enhanced Queuing DPC differs from the VLAN (Level3) shaping on a 1-Gigabit Ethernet Enhanced Queuing DPC. To use the VLAN (Level 3) shaping on a 10-Gigabit Ethernet MX Series Enhanced Queuing DPC, configure an interface set at the [edit interfaces interface-set] hierarchy level. The interface set configuration is not required for configuring a 1-Gigabit Ethernet VLANs on the same Enhanced Queuing DPC.

The Enhanced Queuing DPC supports the following features for scalability:

  • 16,000 queues per Packet Forwarding Engine

  • 4 Packet Forwarding Engines per DPC

    • 4000 schedulers at logical interface level (Level 3) with 4 queues each

    • 2000 schedulers at logical interface level (Level 3) with 8 queues each

  • 255 schedulers at the interface set level (Level 2) per 1-port Packet Forwarding Engine on a 10-Gigabit Ethernet DPC

  • 15 schedulers at the interface set level (Level 2) per 10-port Packet Forwarding Engine on a 1-Gigabit Ethernet DPC

  • About 400 milliseconds of buffer delay (this varies by packet size and if large buffers are enabled)

  • 4 levels of priority (strict-high, high, medium, and low)

Note:

Including the transmit-rate rate exact statement at the [edit class-of-service schedulers scheduler-name] hierarchy level is not supported on Enhanced Queuing DPCs on MX Series routers.

The way that the Enhanced Queuing DPC maps a queue to a scheduler depends on whether 8 queues or 4 queues are configured. By default, a scheduler at level 3 has 4 queues. Level 3 scheduler X controls queue X*4 to X*4+3, so that scheduler 100 (for example) controls queues 400 to 403. However, when 8 queues per scheduler are enabled, the odd numbered schedulers are disabled, allowing twice the number of queues per subscriber as before. With 8 queues, level 3 scheduler X controls queue X*4 to X*4+7, so that scheduler 100 (for example) now controls queues 400 to 407.

You configure the max-queues-per-interface statement to set the number of queues at 4 or 8 at the FPC level of the hierarchy. Changing this statement results in a restart of the DPC. For more information about the max-queues-per-interface statement, see the Junos OS Network Interfaces Library for Routing Devices.

The Enhanced Queuing DPC maps level 3 (customer VLAN) schedulers in groups to level 2 (service VLAN) schedulers. Sixteen contiguous level 3 schedulers are mapped to level 2 when 4 queues are enabled, and 8 contiguous level 3 schedulers are mapped to level 2 when 8 queues are enabled. All of the schedulers in the group should use the same queue priority mapping. For example, if the queue priorities of one scheduler are high, medium, low, and low, then all members of this group should have the same queue priority.

Mapping of a group at level 3 to level 2 can be done at any time. However, a group at level 3 can only be unmapped from a level 2 scheduler only if all the schedulers in the group are free. Once unmapped, a level 3 group can be remapped to any level 2 scheduler. There is no restriction on the number of level 3 groups that can be mapped to a particular level 2 scheduler. There can be 256 level 3 groups, but fragmentation of the scheduler space can reduce the number of schedulers available. In other words, there are scheduler allocation patterns that might fail even though there are free schedulers.

In contrast to level-3-to-level-2 mapping, the Enhanced Queuing DPC maps level 2 (service VLAN) schedulers in a fixed mode to level 1 (physical interface) schedulers. On 40-port Gigabit Ethernet DPCs, there are 16 level 1 schedulers, and 10 of these are used for the physical interfaces. There are 256 level 2 schedulers, or 16 per level 1 scheduler. A level 1 scheduler uses level schedulers X*16 through X*16+15. So level 1 scheduler 0 uses level 2 schedulers 0 through 15, level 1 scheduler 1 uses level 2 schedulers 16 through 31, and so on. On 4-port 10-Gigabit Ethernet PICs, there is one level 1 scheduler for the physical interface, and 256 level 2 schedulers are mapped to the single level 1 scheduler.

The maximum number of level 3 (customer VLAN) schedulers that can be used is 4076 (4 queues) or 2028 (8 queues) for the 10-port Gigabit Ethernet Packet Forwarding Engine and 4094 (4 queues) or 2046 (8 queues) for the 10-Gigabit Ethernet Packet Forwarding Engine.

Enhanced Queuing is supported on aggregated Ethernet (AE) interfaces with two links in link protection mode. However, only one link in the AE bundle can be active at a time. Traffic is shaped independently on the two links, but the member’s links do not need to reside in the same Packet Forwarding Engine or the same DPC. Finally, shared schedulers are not supported on the Enhanced Queuing DPC (use hierarchical schedulers to group logical interfaces).