US 11,870,705 B1
De-scheduler filtering system to minimize service disruptions within a network
Roy Nidesh Kantharajah, La Crosse, WI (US); Wojciech Dec, Amsterdam (NL); and Krishna Kumar Kuruvadi, San Jose, CA (US)
Assigned to CISCO TECHNOLOGY, INC., San Jose, CA (US)
Filed by Cisco Technology, Inc., San Jose, CA (US)
Filed on Jul. 1, 2022, as Appl. No. 17/856,820.
Int. Cl. H04L 47/762 (2022.01); H04L 47/125 (2022.01); H04L 47/80 (2022.01); H04L 47/70 (2022.01)
CPC H04L 47/762 (2013.01) [H04L 47/125 (2013.01); H04L 47/805 (2013.01); H04L 47/822 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
identifying a node within a cluster of nodes with resources that are at least one of underutilized or overutilized;
identifying pods associated with the cluster of nodes that are eligible to be moved, each of the pods running respective virtual containers and supporting a single service;
generating disruption scores for the pods, each of the disruption scores indicating a measure of disruption caused to the service based on the respective pod being moved;
selecting, based at least in part on a comparison of the disruption scores, a pod from the pods having a low disruptive score relative to the disruption scores;
determining to move the pod based at least in part on the pod having the low disruptive score relative to the disruption scores of the pods; and
sending, to at least one of the cluster of nodes, an instruction to move the pod having the low disruptive score.