System and methods for configuring, deploying and maintaining computing clusters

A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, the computer-program product including instructions operable to cause at least one processor to conditionally repeat configuration of a computing cluster, wherein:

• the computing cluster employs a combination of at least a selected level of hosting service and a selected resource allocation routine;
• an initial set of configuration settings for the computing cluster is generated during an initial configuration of the computing cluster performed at a time prior to operation of the computing cluster;
• the initial configuration of the computing cluster is based on at least one prediction generated using a pre-existing cluster model that is derived from observations of operation of at least one pre-existing computing cluster;
• the at least one pre-existing computing cluster employs a combination of at least a level of hosting service and a resource allocation routine that is identical to the combination of at least the selected level of hosting service and the selected resource allocation routine of the computing cluster;
• at least one node device provides the selected level of hosting service to host the computing cluster;
• the selected level of hosting service defines a level of processing resource and a level of storage resource provided by the at least one node device to support execution of a selected resource allocation routine;
• the selected resource allocation routine allocates the processing resource and the storage resource of the at least one node device through provision of multiple containers;
• each container of the multiple containers provides an execution environment in which to execute instructions of an executable routine to perform a task;
• and
• the at least one processor is caused to perform operations comprising:
• use at least one of a level of resource observed to be consumed by operation of the computing cluster or a level of performance observed to be provided by operation of the computing cluster as an input to the pre-existing cluster model to derive one of a predicted level of performance to be provided by operation of the computing cluster or of a level of resource to be consumed by operation of the computing cluster;
• compare the predicted level to a corresponding observed level of resource consumed or of performance provided;
• and
• in response to the predicted level not matching the observed level to within a pre-selected degree, perform operations comprising:
• derive a new cluster model from observations of the operation of the computing cluster;
• and
• generate a prompt to perform a repetition of the configuration of the computing cluster using the new cluster model in place of the pre-existing cluster model to generate a new set of configuration settings for the computing cluster that replaces the initial set of configuration settings.

A computer-implemented method for conditionally repeating configuration of a computing cluster, wherein:

• the computing cluster employs a combination of at least a selected level of hosting service and a selected resource allocation routine;
• an initial set of configuration settings for the computing cluster is generated during an initial configuration of the computing cluster performed at a time prior to operation of the computing cluster;
• the initial configuration of the computing cluster is based on at least one prediction generated using a pre-existing cluster model that is derived from observations of operation of at least one pre-existing computing cluster;
• the at least one pre-existing computing cluster employs a combination of at least a level of hosting service and a resource allocation routine that is identical to the combination of at least the selected level of hosting service and the selected resource allocation routine of the computing cluster;
• at least one node device provides the selected level of hosting service to host the computing cluster;
• the selected level of hosting service defines a level of processing resource and a level of storage resource provided by the at least one node device to support execution of a selected resource allocation routine;
• the selected resource allocation routine allocates the processing resource and the storage resource of the at least one node device through provision of multiple containers;
• each container of the multiple containers provides an execution environment in which to execute instructions of an executable routine to perform a task;
• and
• the method comprises:
• using, by at least one processor, at least one of a level of resource observed to be consumed by operation of the computing cluster or a level of performance observed to be provided by operation of the computing cluster as an input to the pre-existing cluster model to derive one of a predicted level of performance to be provided by operation of the computing cluster or of a level of resource to be consumed by operation of the computing cluster;
• comparing, by the at least one processor, the predicted level to a corresponding observed level of resource consumed or of performance provided;
• and
• in response to the predicted level not matching the observed level to within a pre-selected degree, performing operations comprising:
• deriving, by the at least one processor, a new cluster model from observations of the operation of the computing cluster;
• and
• generating, by the at least one processor, a prompt to perform a repetition of the configuration of the computing cluster using the new cluster model in place of the pre-existing cluster model to generate a new set of configuration settings for the computing cluster that replaces the initial set of configuration settings.