ENSURING PREDICTABLE AND QUANTIFIABLE NETWORKING PERFORMANCE

• US 20130343191A1
• Filed: 06/21/2012
• Published: 12/26/2013
• Est. Priority Date: 06/21/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Kim, Changhoon, Greenberg, Albert G., Dabagh, Alireza, Khalidi, Yousef A., Bansal, Deepak, Kandula, Srikanth
• Granted Patent: US 8,804,523 B2
• Time in Patent Office: Days
• US Class Current: 370/235
• CPC Class Codes: G06F 2009/45595 Network integration; Enabling network access in virtual machine instances
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: H04L 47/11 Identifying congestion
• CPC Class Codes: H04L 47/12 Avoiding congestion; Recovering from congestion
• CPC Class Codes: H04L 47/263 Rate modification at the source after receiving feedback
• CPC Class Codes: H04L 47/762 triggered by the network
• CPC Class Codes: H04L 47/822 Collecting or measuring resource availability data
• CPC Class Codes: H04L 49/70 Virtual switches
• CPC Class Codes: H04L 69/16 Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
• CPC Class Codes: Y02D 30/50 in wire-line communication networks, e.g. low power modes or reduced link rate

Abstract

The ensuring of predictable and quantifiable networking performance. Embodiments of the invention combine a congestion free network core with a hypervisor based (i.e., edge-based) throttling design to help insure quantitative and invariable subscription bandwidth rates. A lightweight shim layer in a hypervisor can adaptively throttle the rate of VM-to-VM traffic flow. A receiving hypervisor can detect congestion and communicate back to sending hypervisors that rates are to be regulated. In response, sending hypervisors can reduce transmission rate to mitigate congestion at the receiving hypervisor. In some embodiments, the principles are extended to any message processors communicating over a congestion free network.

Claims

1. At a computer system including one or more processors and system memory, the computer system connected to a congestion free network core, the computer system also including a message processor for processing data flows received from the congestion free network core, the message processor associated with a congestion detector, a method for addressing network congestion at the computer system, the method comprising:

an act of accessing a subscription bandwidth for the message processor, the subscription bandwidth indicating a quantitative and invariant minimum bandwidth for the message processor;

an act of receiving one or more data flows from the congestion free network core, the one or more data flows sent from sending message processors and directed to the message processor;

an act of calculating the combined bandwidth of the one or more data flows;

an act of detecting the onset of congestion at the computing system;

an act of identifying at least one message processor associated with the one or more data flows as a violator of the subscription bandwidth in response to detecting the onset of congestion, the at least one violating message processor being a sending message processor or a receiving message processor of one of the one or more data flows;

an act of determining the extent of the violation by the at least one violating message processor;

an act of formulating feedback for delivery to sender side adaptive rate limiters that serves to rate limit the at least one violating message processor, the feedback instructing the sender side adaptive rate limiters to reduce the bandwidth of the one or more data flows originating from the at least one violating message processor; and

an act of sending the feedback onto the congestion free network core for delivery to the sender side adaptive rate limiters.

2. The method as recited in claim 1, wherein the act of receiving one or more data flows from the congestion free network core comprises:

an act of receiving at least one data flow of Transmission Control Protocol (TCP) traffic; and

an act of receiving at least one other data flow of traffic using another different protocol, the other different protocol different than Transmission Control Protocol (TCP).

3. The method as recited in claim 2, wherein the act of identifying at least one message processor associated with the one or more data flows as a violator of the subscription bandwidth comprises an act of identifying a specified message processor associated with the at least one other data flow as a violator of the subscription bandwidth.

4. The method as recited in claim 3, wherein the act of formulating feedback for delivery to sender side adaptive rate limiters corresponding to the at least one violating message processor comprises an act of formulating feedback for delivery to a sender side rate limiter that serves to rate limit the specified message processor, the feedback instructing the sender side adaptive rate limiter to reduce the bandwidth of the at least one other data flow.

5. The method as recited in claim 4, wherein the act of formulating feedback for delivery to a sender side rate limiter the serves to rate limit the specified message processor comprises an act of formulating feedback for reducing the bandwidth of the at least one other data flow in accordance with an additive increase/multiplicative decrease (“

AIMD”

) algorithm such that the traffic using the other different protocol is reduced based on the additive increase/multiplicative decrease (“

AIMD”

) algorithm.

6. The method as recited in claim 2, wherein the act of receiving at least one other data flow of traffic using another different protocol comprises an act of receiving at least one data flow of User Datagram Protocol (“

UDP”

) traffic.

7. The method as recited in claim 1, wherein the act of detecting the onset of congestion at the computing system comprises an act of a software meter detecting the onset of congestion at the computing system.

8. The method as recited in claim 1, wherein the act of detecting the onset of congestion at the computing system comprises an act of using Explicit Congestion Notification (“

ECN”

) to detect the onset of congestion at the computing system.

9. The method as recited in claim 1, wherein the message processor is a virtual machine served by a hypervisor, and wherein the act of detecting the onset of congestion is performed by a virtual switch within the hypervisor.

10. The method as recited in claim 1, wherein the at least one violating message processor includes at least one virtual machine that is served by a hypervisor, wherein the adaptive rate limiter that servers to rate limit the virtual machine is included in a virtual switch of the hypervisor.

11. The method as recited in claim 1, wherein the act of formulating feedback for delivery to sender side adaptive rate limiters corresponding to the at least one violating message processor comprises an act of formulating feedback for reducing the bandwidth of the one or more data flows originating from the at least one violating message processor in accordance with an additive increase/multiplicative decrease (“

AIMD”

) algorithm.

12. A computer program product for use at a computer system, the computer system connected to a congestion free network core, the computer system also including a message processor for processing data flows received from the congestion free network core, the message processor associated with a congestion detector, the computer program product for implementing a method for addressing network congestion at the computer system, the computer program product claim comprising one or more computer storage devices having stored thereon computer-executable instructions that, when executed at a processor, cause the computer system to perform the method, including the following:

access a subscription bandwidth for the message processor, the subscription bandwidth indicating a quantitative and invariant minimum bandwidth for the message processor;

receive one or more data flows from the congestion free network core, the one or more data flows sent from sending message processor and directed to the message processor;

calculate the combined bandwidth of the one or more data flows;

detect the onset of congestion at the computing system;

identify at least one message processor associated with the one or more data flows as a violator of the subscription bandwidth in response to detecting the onset of congestion, the at least one violating message processor being a sending message processor or a receiving message processor of one of the one or more data flows;

determine the extent of the violation by the at least one violating message processor;

formulate feedback for delivery to sender side adaptive rate limiters that serve to rate limit the at least one violating message processor, the feedback instructing the sender side adaptive rate limiters to reduce the bandwidth of the one or more data flows originating from the at least one violating message processor; and

send the feedback onto the congestion free network core for delivery to the sender side adaptive rate limiters.

13. The computer program product as recited in claim 12, wherein computer-executable instructions that, when executed, cause the computer system to receive one or more data flows from the congestion free network core comprise computer-executable instructions that, when executed, cause the computer system to:

receive at least one data flow of Transmission Control Protocol (TCP) traffic; and

receive at least one other data flow of traffic using another different protocol, the other different protocol different than Transmission Control Protocol (TCP).

14. The computer program product as recited in claim 12, wherein computer-executable instructions that, when executed, cause the computer system to formulate feedback for delivery to sender side adaptive rate limiters corresponding to the at least one violating message processor comprise computer-executable instructions that, when executed, cause the computer system to formulate feedback for reducing data flow bandwidth in accordance with an additive increase/multiplicative (“

AIMD”

) decrease algorithm.

15. The computer program product as recited in claim 12, wherein computer-executable instructions that, when executed, cause the computer system to detect the onset of congestion at the computing system comprise computer-executable instructions that, when executed, cause the computer system to use software metering to detect the onset of congestion at the computing system.

16. The computer program product as recited in claim 12, wherein computer-executable instructions that, when executed, cause the computer system to detect the onset of congestion at the computing system comprise computer-executable instructions that, when executed, cause the computer system to use Explicit Congestion Notification (“

ECN”

) to detect the onset of congestion at the computing system.

17. The computer program product as recited in claim 12, wherein the message processor is a virtual machine served by a hypervisor, and wherein the act of detecting the onset of congestion is performed by a virtual switch within the hypervisor.

18. A system, the system comprising:

a congestion free network core; and

a plurality of hypervisors connected to the congestion free network core, each of the plurality of hypervisors running on a hardware host, each of the plurality of hypervisors including a virtual machine switch and managing one or more virtual machines, each virtual machine switch including at least one of a congestion detector and an adaptive rate limiter;

each congestion detection configured to;

access a subscription bandwidth for a virtual machine managed by a corresponding hypervisor, the subscription bandwidth indicating a quantitative and invariant minimum bandwidth for the virtual machine;

receive one or more data flows from the congestion free network core, the one or more data flows sent from other virtual machines and directed to the virtual machine;

calculate the combined bandwidth of the one or more data flows;

detect the onset of congestion at the virtual machine switch;

identify at least one of the other virtual machines associated with the one or more data flows as a violator of the subscription bandwidth in response to detecting the onset of congestion;

determine the extent of the violation by the at least one violating virtual machine;

formulate feedback for delivery to an adaptive rate limiters at the virtual machine switch for the at least one violating virtual machine, the feedback instructing the adaptive rate limiters to reduce the bandwidth of the one or more data flows originating from the at least one violating virtual machine; and

send the feedback onto the congestion free network core for delivery to the adaptive rate limiters at one or more virtual machine switches; and

each adaptive rate limiter configured to;

receive feedback from the congestion free network core, the feedback sent from congestion detectors at a second one or more virtual machine switches; and

reduce the rate of one or more data flows in accordance with a congestion avoidance algorithm in response to receiving the feedback.

19. The system of claim 18, wherein each congestion detector configured to detect the onset of congestion at the virtual machine switch comprises each congestion detector configured to detect the onset of congestion using one or more of software metering and Explicit Congestion Notification (“

ECN”

).

20. The system of claim 18, wherein each adaptive rate limiter configured to reduce the rate of one or more data flows in accordance with a congestion avoidance algorithm in response to receiving the feedback comprises each adaptive rate limiter configured to reduce the rate of User Datagram Protocols (“

UDP”

) data flows in accordance with an additive increase/multiplicative (“

AIMD”

) decrease algorithm.