US 12,069,128 B2
Automatic scaling of resource instance groups within compute clusters
Jonathan Daly Einkauf, Seattle, WA (US); Luca Natali, Kirkland, WA (US); Bhargava Ram Kalathuru, Seattle, WA (US); Saurabh Dileep Baji, Seattle, WA (US); and Abhishek Rajnikant Sinha, Bellevue, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Jun. 18, 2021, as Appl. No. 17/352,065.
Application 17/352,065 is a continuation of application No. 16/805,412, filed on Feb. 28, 2020, granted, now 11,044,310.
Application 16/805,412 is a continuation of application No. 15/845,855, filed on Dec. 18, 2017, granted, now 10,581,964, issued on Mar. 3, 2020.
Application 15/845,855 is a continuation of application No. 14/702,080, filed on May 1, 2015, granted, now 9,848,041, issued on Dec. 19, 2017.
Prior Publication US 2021/0392185 A1, Dec. 16, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. H04L 41/0893 (2022.01); G06F 9/50 (2006.01); H04L 41/0894 (2022.01); H04L 41/0897 (2022.01); H04L 41/22 (2022.01); H04L 41/5041 (2022.01); H04L 43/0876 (2022.01); H04L 67/10 (2022.01); H04L 67/1031 (2022.01); H04L 67/1074 (2022.01)
CPC H04L 67/1076 (2013.01) [G06F 9/5077 (2013.01); G06F 9/5083 (2013.01); H04L 41/0893 (2013.01); H04L 41/0894 (2022.05); H04L 41/0897 (2022.05); H04L 41/22 (2013.01); H04L 41/5045 (2013.01); H04L 43/0876 (2013.01); H04L 67/10 (2013.01); H04L 67/1031 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
performing, by one or more computers:
implementing an interface configured to provide for selection of a type of condition from among a plurality of different types of conditions available for an automatic scaling policy associated with a cluster of nodes comprising two or more non-overlapping instance groups, each instance group comprising a respective one or more computing resource instances, the types of conditions available for selection comprising a workload-based condition and a time-based condition;
receiving, via the interface, one or more inputs from a client:
selecting from among the workload-based condition or the time-based condition for the automatic scaling policy, and
associating the automatic scaling policy that specifies the selected condition with the cluster of nodes;
detecting that a trigger condition, comprising the workload-based condition or the time-based condition, specified in the automatic scaling policy, has been met during execution of a distributed application on the cluster of nodes; and
in response to said detecting, performing an automatic scaling operation comprising one of a workload-based auto-scale or a time-based auto-scale, as specified by the corresponding condition specified in the automatic scaling policy, on one of the two or more non-overlapping instance groups, the one of the two or more non-overlapping instance groups specified by the automatic scaling policy, wherein the automatic scaling operation changes a number of computing resource instances on the one of the two or more non-overlapping instance groups.