US 12,436,799 B2
Cold-start service placement over on-demand resources
Venkatesan Thirumalai Chakaravarthy, Bangalore (IN); Ashok Pon Kumar Sree Prakash, Bangalore (IN); Saritha Vinod, Bangalore (IN); and Yogish Sabharwal, Gurgaon (IN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Dec. 15, 2021, as Appl. No. 17/551,520.
Prior Publication US 2023/0185604 A1, Jun. 15, 2023
Int. Cl. G06F 9/48 (2006.01); G06F 9/50 (2006.01); H04L 67/10 (2022.01)
CPC G06F 9/4881 (2013.01) [G06F 9/5044 (2013.01); H04L 67/10 (2013.01); G06F 2209/501 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method, the method comprising:
obtaining a profile comprising performance requirements of a service that vary over time;
determining a plurality of service incarnations for the service, wherein each service incarnation is associated with a distinct level of performance provided by the service incarnation for the service, distinct resource requirements of the service incarnation, and a computing node type the service incarnation is configured to execute on;
identifying a plurality of computing nodes having different computing node types and different resource capacities;
jointly scheduling the plurality of computing nodes and one or more of the plurality of service incarnations on the plurality of computing nodes over a time interval such that a cumulative level of performance of the service incarnations scheduled at each timepoint in the time interval satisfies the profile of the service, wherein the scheduling is based at least in part on the resource requirements of the plurality of service incarnations and the resource capacities of each of the computing nodes; and
triggering execution of the service in accordance with the scheduling;
wherein the method is carried out by at least one computing device.