CPC G06F 11/2028 (2013.01) [G06F 11/008 (2013.01); G06F 11/2025 (2013.01)] | 17 Claims |
1. A system, comprising:
one or more processors and memory that includes instructions that, when executed by the one or more processors, cause the one or more processors to:
monitor a health characteristic of a first region designated as a primary region of a non-relational database cluster that provides data services to an application, wherein the non-relational database cluster includes a plurality of geographically separate regions and employs majority election of a primary in automatic failover;
invoke a machine learning model on the health characteristic, wherein the machine learning model is trained on historical failover data to predict a likelihood of a failover state of the primary region; and
designate a second region of the plurality of geographically separate regions as the primary region to provide the data services to the application when the likelihood of the failover state satisfies a predetermined threshold, wherein the instructions specify a lambda function that designates the second region as the primary region.
|