POWER AWARE SWITCHING USING ANALYTICS

• US 20180074570A1
• Filed: 09/09/2016
• Published: 03/15/2018
• Est. Priority Date: 09/09/2016
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Akella, Anand V., Iyengar, Praveen Parthasarathy, Thirumurthi, Rajendra Kumar, Sharma, Samar, Dharwada, Krishna Bharadwaj, Purushotham, Vivek
• Granted Patent: US 10,209,763 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 1/3215 Monitoring of peripheral devices
• CPC Class Codes: G06F 1/3278 Power saving in modem or I/O interface
• CPC Class Codes: G06F 1/3287 by switching off individual functional units in the computer system
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: H04L 41/0833 for reduction of network energy consumption
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04L 41/149 for prediction of maintenance
• CPC Class Codes: H04L 41/16 using machine learning or artificial intelligence
• CPC Class Codes: H04L 43/08 Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
• CPC Class Codes: H04L 43/0876 Network utilisation, e.g. volume of load or congestion level
• CPC Class Codes: H04L 45/08 Learning-based routing, e.g. using neural networks or artificial intelligence
• CPC Class Codes: H04L 45/12 Shortest path evaluation
• CPC Class Codes: H04L 45/123 Evaluation of link metrics techniques for monitoring network metrics H04L43/08
• CPC Class Codes: H04L 45/70 Routing based on monitoring results
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management
• CPC Class Codes: Y02D 30/50 in wire-line communication networks, e.g. low power modes or reduced link rate

Abstract

A method is described and in one embodiment includes, for each of a plurality of outgoing ports of a first network element: collecting data comprising a number of packets arriving the outgoing port and an amount of power consumed by the outgoing port for a first time interval; calculating a packet per watt (“P/W”) metric for the port for the first time interval, wherein the P/W metric comprises the number of packets coming into the port divided by the amount of power consumed by the port during the first time interval; repeating the collecting and calculating for a number of successive time intervals; calculating a mean P/W metric for a time period comprising the first time interval and the successive time intervals; and calculating a variance for the time period comprising the first time interval and the successive time intervals. The method further includes redirecting traffic received at the network element to the outgoing port having the lowest variance.

Claims

1. A method comprising:

for each of a plurality of outgoing ports of a first network element;

collecting data comprising a number of packets arriving the outgoing port and an amount of power consumed by the outgoing port for a first time interval;

calculating a packet per watt (“

P/W”

) metric for the port for the first time interval, wherein the P/W metric comprises the number of packets coming into the port divided by the amount of power consumed by the port during the first time interval;

repeating the collecting and calculating for a number of successive time intervals;

calculating a mean P/W metric for a time period comprising the first time interval and the successive time intervals; and

calculating a variance for the time period comprising the first time interval and the successive time intervals; and

redirecting traffic received at the network element to the outgoing port having the lowest variance.

2. The method of claim 1, wherein the redirecting comprises assigning the calculated variance to a link associated with the outgoing port as a cost metric of the link for use by a link state protocol.

3. The method of claim 1 further comprising, for each of the outgoing ports, calculating a standard deviation for the time period comprising the first time interval and the successive time intervals.

4. The method of claim 1 further comprising storing the collected data for each of the outgoing ports in a central database associated with the network element.

5. The method of claim 4 further comprising:

processing newly collected data against the stored data using a machine learning algorithm; and

initiating a network power-saving action based on the processing.

6. The method of claim 5 wherein the power-saving action comprises one of displaying per-port power savings for each of the outgoing ports, powering down one of the outgoing ports and moving traffic to another one of the outgoing ports;

powering down a portion of the network element;

powering down a line card of the network element on which at least one of the outgoing ports is disposed; and

powering down the network element in its entirety.

7. The method of claim 1, wherein the link state protocol comprises one of Open Shortest Path First (“

OSPF”

) and Intermediate System to Intermediate System (“

ISIS”

).

8. The method of claim 1, wherein the network element comprises a leaf switch in a fully meshed spine-leaf network.

9. One or more non-transitory tangible media that includes code for execution and when executed by a processor is operable to perform operations comprising:

for each of a plurality of outgoing ports of a first network element;

collecting data comprising a number of packets arriving the outgoing port and an amount of power consumed by the outgoing port for a first time interval;

calculating a packet per watt (“

P/W”

) metric for the port for the first time interval, wherein the P/W metric comprises the number of packets coming into the port divided by the amount of power consumed by the port during the first time interval;

repeating the collecting and calculating for a number of successive time intervals;

calculating a mean P/W metric for a time period comprising the first time interval and the successive time intervals; and

calculating a variance for the time period comprising the first time interval and the successive time intervals; and

redirecting traffic received at the network element to the outgoing port having the lowest variance.

10. The media of claim 9, wherein the redirecting comprises assigning the calculated variance to a link associated with the outgoing port as a cost metric of the link for use by a link state protocol.

11. The media of claim 9, wherein the operations further comprise:

storing the collected data for each of the outgoing ports in a central database associated with the network element;

processing newly collected data against the stored data using a machine learning algorithm; and

initiating a network power-saving action based on the processing.

12. The media of claim 11, wherein the power-saving action comprises one of displaying per-port power savings for each of the outgoing ports, powering down one of the outgoing ports and moving traffic to another one of the outgoing ports;

powering down a portion of the network element;

powering down a line card of the network element on which at least one of the outgoing ports is disposed; and

powering down the network element in its entirety.

13. The media of claim 9, wherein the link state protocol comprises one of Open Shortest Path First (“

OSPF”

) and Intermediate System to Intermediate System (“

ISIS”

).

14. The media of claim 9, wherein the network element comprises a leaf switch in a fully meshed spine-leaf network.

15. An apparatus comprising:

a memory element configured to store data;

a processor operable to execute instructions associated with the data; and

a power savings module configured to;

for each of a plurality of outgoing ports of a first network element;

collect data comprising a number of packets arriving the outgoing port and an amount of power consumed by the outgoing port for a first time interval;

calculate a packet per watt (“

P/W”

) metric for the port for the first time interval, wherein the P/W metric comprises the number of packets coming into the port divided by the amount of power consumed by the port during the first time interval;

repeat the collecting and calculating for a number of successive time intervals;

calculate a mean P/W metric for a time period comprising the first time interval and the successive time intervals; and

calculate a variance for the time period comprising the first time interval and the successive time intervals; and

redirect traffic received at the network element to the outgoing port having the lowest variance.

16. The apparatus of claim 15, wherein the redirecting comprises assigning the calculated variance to a link associated with the outgoing port as a cost metric of the link for use by a link state protocol.

17. The apparatus of claim 15, wherein the apparatus is further configured to:

store the collected data for each of the outgoing ports in a central database associated with the network element;

process newly collected data against the stored data using a machine learning algorithm; and

initiate a network power-saving action based on the processing.

18. The apparatus of claim 17, wherein the power-saving action comprises one of displaying per-port power savings for each of the outgoing ports, powering down one of the outgoing ports and moving traffic to another one of the outgoing ports;

powering down a portion of the network element;

powering down a line card of the network element on which at least one of the outgoing ports is disposed; and

powering down the network element in its entirety.

19. The apparatus of claim 15, wherein the link state protocol comprises one of Open Shortest Path First (“

OSPF”

) and Intermediate System to Intermediate System (“

ISIS”

).

20. The apparatus of claim 15, wherein the network element comprises a leaf switch in a fully meshed spine-leaf network.