Coordinated queuing between upstream and downstream queues in a network device

• US 7,899,068 B1
• Filed: 10/09/2007
• Issued: 03/01/2011
• Est. Priority Date: 10/09/2007
• Status: Active Grant

• Current Assignee: Juniper Networks Incorporated
• Original Assignee: Juniper Networks Incorporated
• Inventors: Su, Jiaxiang, Ma, Qingming
• Primary Examiner(s): Ton; Dang T
• Assistant Examiner(s): KAMARA, MOHAMED A
• Application Number:
  US11/869,327
• Time in Patent Office: 1,239 Days
• Field of Search: 370/235, 370/392, 370/395.4, 370/412, 370/468, 370/474, 710/39
• US Class Current: 370/412
• CPC Class Codes: H04L 47/6255 queue load conditions, e.g. longest queue first
• CPC Class Codes: H04L 47/6275 based on priority
• CPC Class Codes: H04L 49/254 Centralised controller, i.e. arbitration or scheduling
• CPC Class Codes: H04L 49/30 Peripheral units, e.g. input or output ports
• CPC Class Codes: H04L 49/90 Buffering arrangements

Abstract

A system determines a scheduling value based on a current length of a downstream queue in a network device. The system sends the scheduling value from the downstream queue to an upstream queue and schedules dequeuing of one or more data units, destined for the downstream queue, from the upstream queue based on the scheduling value.

Claims

1. A network device, comprising:

a first ingress card to receive a first data unit from a network and including a first upstream queue that queues the received first data unit;

a switch fabric coupled to the ingress card and to a first egress card and to transfer data units from the ingress card to the first egress card;

the first egress card including a first downstream queue for buffering the first data unit prior to forwarding the first data unit towards a destination via the network, the first egress card to;

determine a first scheduling value as a function of a priority of the first downstream queue, a maximum size of the first downstream queue, and a current length of the first downstream queue, and

send the first scheduling value to the first ingress card,

where the first ingress card is further to schedule dequeuing of the first data unit from the first upstream queue based on the first scheduling value,

where the first ingress card is further to receive a second data unit from the network and includes a second upstream queue that queues the received second data unit and further comprising;

a second egress card that includes a second downstream queue for buffering the second data unit prior to forwarding the second data unit towards a destination via the network, the second egress card to;

determine a second scheduling value as a function of a priority and current length of the second downstream queue, and

send the second scheduling value to the first ingress card; and

where the first ingress card is further to schedule dequeuing of the second data unit from the second upstream queue based on the second scheduling value.

2. The network device of claim 1, further comprising:

a second ingress card to receive a second data unit from the network and including a second upstream queue that queues the received second data unit;

the egress card to send the first scheduling value to the second ingress card;

where the second ingress card is further to schedule dequeueing of the second data unit from the second upstream queue based on the first scheduling value.

3. The network device of claim 1, where the function comprises:

4. A method, comprising:

receiving, by a first ingress card in a network device, a first data unit from a network, the first ingress card including a first upstream queue that queues the received first data unit;

transferring, by a switch fabric coupled to the ingress card, data units from the ingress card to the first egress card;

buffering, by the first egress card, the first data unit, in a first downstream queue, prior to forwarding the first data unit towards a destination via the network, including;

determining a first scheduling value as a function of a priority of the first downstream queue, a maximum size of the first downstream queue, and a current length of the first downstream queue, and

sending the first scheduling value to the first ingress card;

scheduling, by the first ingress card, dequeuing of the first data unit from the first upstream queue based on the first scheduling value;

receiving, by the first ingress card, a second data unit from the network, the first ingress card further including a second upstream queue that queues the received second data unit;

buffering, by a second egress card, the second data unit, in a second downstream queue, prior to forwarding the second data unit towards a destination via the network, including;

determining a second scheduling value as a function of a priority and current length of the second downstream queue, and

sending the second scheduling value to the first ingress card; and

scheduling, by the first ingress card, dequeuing of the second data unit from the second upstream queue based on the second scheduling value.

6. The method of claim 4, where the function comprises:

5. The method of 4, further comprising:

a second ingress card to receive a second data unit from the network and including a second upstream queue that queues the received second data unit;

the egress card to send the first scheduling value to the second ingress card;

where the second ingress card is further to schedule dequeueing of the second data unit from the second upstream queue based on the first scheduling value.