US 12,135,998 B2
Managing computational bursting on server nodes
Ioannis Manousakis, Redmond, WA (US); Husam Alissa, Redmond, WA (US); Ashish Raniwala, Bellevue, WA (US); Ricardo Gouvêa Bianchini, Bellevue, WA (US); Pulkit Ambikanandan Misra, Bellevue, WA (US); Inigo Goiri Presa, Redmond, WA (US); and Majid Jalili, Austin, TX (US)
Assigned to Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on Feb. 26, 2021, as Appl. No. 17/186,856.
Prior Publication US 2022/0276905 A1, Sep. 1, 2022
Int. Cl. G06F 9/50 (2006.01); G06F 9/455 (2018.01); H04L 43/10 (2022.01); H04L 47/70 (2022.01); H05K 7/20 (2006.01)
CPC G06F 9/5077 (2013.01) [G06F 9/45558 (2013.01); H04L 43/10 (2013.01); H04L 47/82 (2013.01); H05K 7/203 (2013.01); H05K 7/20809 (2013.01); G06F 2009/4557 (2013.01); G06F 2209/501 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving burst notifications from a plurality of server nodes of a computing zone, the burst notifications including computational bursting data from the plurality of server nodes, wherein the computational bursting data is tracked and maintained locally by each of the plurality of server nodes and transmitted by each of the plurality of server nodes in response to each of the plurality of server nodes engaging in computational bursting at one or more levels;
maintaining a computing zone record including the computational bursting data for the plurality of server nodes of the computing zone;
determining, based on the computational bursting data, a lifetime for each of the plurality of server nodes, wherein the lifetime is based on a duration of the computational bursting of each server node at the one or more levels;
receiving a customer request to deploy one or more virtual machines on the computing zone;
identifying one or more server nodes from the plurality of server nodes based on a remaining lifetime of the one or more server nodes; and
causing the one or more virtual machines to be selectively deployed on computing resources of the one or more server nodes in accordance with an allocation policy, wherein the allocation policy includes rules for allocating the one or more virtual machines such that the one or more server nodes engage in a high-bursting workload based on the remaining lifetime of the one or more server nodes.