US 12,093,709 B1
Network performance driven compute workload placement
Fabien Chraim, Wichita Falls, TX (US); Antonio Gomariz Penalver, Dublin (IE); and Irina Lopatina, Dublin (IE)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Dec. 10, 2020, as Appl. No. 17/118,281.
Int. Cl. G06F 9/455 (2018.01); G06F 9/50 (2006.01); G06F 11/30 (2006.01); G06F 11/34 (2006.01)
CPC G06F 9/45558 (2013.01) [G06F 9/5011 (2013.01); G06F 9/5077 (2013.01); G06F 11/3072 (2013.01); G06F 11/3452 (2013.01); G06F 2009/4557 (2013.01); G06F 2009/45595 (2013.01)] 20 Claims
OG exemplary drawing
 
5. A method, comprising:
receiving, at a computing resource management system, a request to obtain one of a plurality of computing resources to host a virtualized resource; and
provisioning the virtualized resource on a particular one of the plurality of computing resources, comprising:
obtaining networking performance for individual ones of the plurality of computing resources according to metrics for the individual ones of the plurality of computing resources, wherein the metrics represent performance of networks through which the individual ones of the plurality of computing resources are accessed, and wherein the metrics include:
first metrics that are calculated based, at least in part, on a measurement of time in compliance that performance of network traffic has with respect to a loss threshold for lost packets sent through the networks to the plurality of computing resources to determine a compliance percentage of a total amount of time in a time window for performing the measurement; and
selecting the particular computing resource out of the plurality of computing resources based, at least in part, on the obtained networking performance for the plurality of computing resources.