Back to results
US 12,706,861 B2Granted
Techniques of handling overlay encapsulation
AssigneeOracle International Corporation
InventorsJagwinder Singh Brar, David Dale Becker
Life of this patent
Estimating the term…
Abstract
A network environment comprises a plurality of host machines that are communicatively coupled to each other via a network fabric comprising a plurality of switches that in turn include a plurality of ports. Each host machine comprises one or more GPUs that execute customer workloads. Described herein are different approaches that provide for addressing the problem of handling network overlay encapsulation without causing adverse impact to the performance of workloads executed on the GPU clusters.
Claims20
A method comprising:
- in a network environment comprising a plurality of host machines that are communicatively coupled to each other via a network fabric comprising a plurality of switches, receiving, by a source host machine, a first packet from a first switch included in the plurality of switches, the first packet indicative of an occurrence of congestion in the network fabric;
- configuring, by the source host machine, in response to receiving the first packet, a network interface card associated with a GPU included in the source host machine, the configuring comprising reducing a transmission rate of the GPU by a first predetermined amount;
- responsive to receiving by the source host machine a second packet from the first switch, wherein the second packet indicates presence of congestion in the network fabric, reconfiguring, by the source host machine, the network interface card associated with the GPU, the reconfiguring comprising further reducing the transmission rate of the GPU by a second predetermined amount;
- and
- responsive to not receiving the second packet, modifying by the source host machine, the transmission rate of the GPU to correspond to an initial transmission rate of the GPU prior to receiving the first packet.
2 more elements
One or more computer readable non-transitory media storing computer-executable instructions that, when executed by one or more processors, cause:
- in a network environment comprising a plurality of host machines that are communicatively coupled to each other via a network fabric comprising a plurality of switches, receiving, by a source host machine, a first packet from a first switch included in the plurality of switches, the first packet indicative of an occurrence of congestion in the network fabric;
- configuring, by the source host machine, in response to receiving the first packet, a network interface card associated with a GPU included in the source host machine, the configuring comprising reducing a transmission rate of the GPU by a first predetermined amount;
- responsive to receiving by the source host machine a second packet from the first switch, wherein the second packet indicates presence of congestion in the network fabric, reconfiguring, by the source host machine, the network interface card associated with the GPU, the reconfiguring comprising further reducing the transmission rate of the GPU by a second predetermined amount;
- and
- responsive to not receiving the second packet, modifying by the source host machine, the transmission rate of the GPU to correspond to an initial transmission rate of the GPU prior to receiving the first packet.
2 more elements
A computing device comprising:
- one or more processors;
- and
- a memory including instructions that, when executed with the one or more processors, cause the computing device to, at least:
- in a network environment comprising a plurality of host machines that are communicatively coupled to each other via a network fabric comprising a plurality of switches, receive, by a source host machine, a first packet from a first switch included in the plurality of switches, the first packet indicative of an occurrence of congestion in the network fabric;
- configure, by the source host machine, in response to receiving the first packet, a network interface card associated with a GPU included in the source host machine, the configuring comprising reducing a transmission rate of the GPU by a first predetermined amount;
- responsive to receiving by the source host machine a second packet from the first switch, wherein the second packet indicates presence of congestion in the network fabric, reconfigure, by the source host machine, the network interface card associated with the GPU by further reducing the transmission rate of the GPU by a second predetermined amount;
- and
- responsive to not receiving the second packet, modify by the source host machine, the transmission rate of the GPU to correspond to an initial transmission rate of the GPU prior to receiving the first packet.
5 more elements
Dependencies are inferred from the claim text.