Self-managed intelligent elastic cloud stack

Patent metadata

Field	Value
Publication number	US-2020004591-A1
Application number	US-201916453206-A
Country	US
Kind code	A1
Filing date	Jun 26, 2019
Priority date	Jun 27, 2018
Publication date	Jan 2, 2020
Grant date	—

Abstract

A cloud resource control circuitry may provide an automatic, self-managed, and a real-time elasticity in an active cloud stack including a software stack. The cloud resource control circuitry particularly adjusts cloud computing resources allocated to the active cloud stack based on real-time remote monitoring of resource utilization of the software stack. The resource adjustment may be performed in a stepwise fashion, in a pipelined manner as to the various layers of a software stack, and using hierarchical approach as to levels of cloud computing resources (e.g., task/level, server level, virtual machine/cloud level), to provide a more efficient cloud computing environment.

First claim

What is claimed is:

1 . A system for real-time control of cloud computing resources allocated to an active cloud stack, comprising:
a repository for storing a first predictive model and cloud stack configuration; and processing circuitry in communication with the repository and with the active cloud stack, the processing circuitry configured to:
identify a first collection of software components of a software stack of the active cloud stack as an upstream layer of the software stack based on the cloud stack configuration;
identify a second collection of software components of the software stack as a middle layer of the software stack based on the cloud stack configuration, wherein the middle layer consumes data generated by the upstream layer; and
iteratively cause adjustment of computing resources allocated to the upstream layer and the middle layer of the software stack by:
communicating, by the processing circuitry, with the active cloud stack to obtain cloud resource utilization parameters for the software stack in real-time;
extracting, by the processing circuitry, a first real-time resource utilization measure of the upstream layer and a second real-time resource utilization measure of the middle layer based on the cloud resource utilization parameters; and
upon determining by the processing circuitry that the first real-time resource utilization measure is outside a first predetermined range:
calculating, by the processing circuitry, a predicted resource utilization of the middle layer based on the first real-time resource utilization measure and using the first predictive model; and
when the second real-time resource utilization measure of the middle layer is unexpectedly higher or lower than the predicted resource utilization of the middle layer, generating by the processing circuitry a first instruction for adjusting the computing resources allocated to the upstream layer and the middle layer by a first predetermined amount, and sending by the processing circuitry the first instruction to the active cloud stack for execution.

2 . The system of claim 1 , wherein the software stack comprises a data analytics stack and wherein the upstream layer comprises a data extraction, transformation, and loading layer.

3 . The system of claim 2 , wherein the middle layer comprises a data quality layer.

4 . The system of claim 1 , wherein the repository further stores a second predictive model and wherein the processing circuitry is further configured to:
identify a third collection of software components of the software stack of the active cloud stack as a downstream layer, wherein the downstream layer consumes data generated by the middle layer; and iteratively cause adjustment of computing resources allocated to the downstream layer by:
extracting, by the processing circuitry, a third real-time resource utilization measure of the downstream layer based on the cloud resource utilization parameters; and
upon determining by the processing circuitry that the third real-time resource utilization measure is outside a second predetermined range:
calculating, by the processing circuitry, a predicted resource utilization of the downstream layer based on at least one of the first real-time resource utilization measure and the second real-time resource utilization measure, and using the second predictive model; and
when the third real-time resource utilization measure of the downstream layer is unexpectedly higher or lower than the predicted resource utilization of the downstream layer, generating by the processing circuitry a second instruction for adjusting the computing resources allocated to the downstream layer by a second predetermined amount, and sending by the processing circuitry the second instruction to the active cloud stack for execution.

5 . The system of claim 4 , wherein the processing circuitry is further configured to identify each of the upstream layer, the middle layer, and the downstream layer of the software stack as belonging to a cluster type or a non-cluster type, wherein:
a layer of the cluster type includes at least one software component running on distributed cloud computing nodes; a layer of the non-cluster type includes software components running on dedicated computing nodes; the cluster type is associated with a first set of cloud adjustment actions and the non-cluster type is associated with a second set of cloud adjustment actions; and the first instruction for adjusting the computing resources allocated to the upstream layer and middle layer and the second instruction for adjusting computing resources allocated to the downstream layer each include adjustment actions selected from the first set of cloud adjustment actions and the second set of cloud adjustment actions based on the identified types for the upstream layer, the middle layer, and the downstream layer.

6 . The system of claim 5 , wherein the first set of cloud adjustment actions encompass the second set of cloud adjustment actions and additionally include a cloud adjustment action for adding or removing a cloud computing node for a distributed software component.

7 . The system of claim 4 , wherein:
the software stack comprises a data analytics stack; the upstream layer of the software stack comprises a data extraction, transformation, and loading layer; the middle layer of the software stack comprises a data quality layer; and the downstream layer of the software stack comprises a data processing or a data analytics layer.

8 . The system of claim 1 , wherein the first instruction for adjusting the computing resources allocated to the upstream layer and the middle layer, during the iterative process, comprises cloud resource adjustment actions that alternate between job level resource allocation adjustment action, server level resource allocation adjustment action, and virtual machine level allocation adjustment action.

9 . The system of claim 1 , wherein the active cloud stack comprises cloud computing resources provisioned by at least two independent cloud provisioning circuitries with distinct credential and access key schemes.

10 . The system of claim 9 , wherein the processing circuitry is further configured to provide a single user interface and a single multi-factor authentication scheme.

11 . The system of claim 1 , wherein the repository further comprises a third predictive model and wherein the processing circuitry is further configured to:
receive a user specification and requisition for deployment of a cloud stack before the active cloud stack was deployed; automatically determine a predicted software stack and a cloud resource configuration based on the user specification and using the third predictive model; and automatically deploy the active cloud stack according to the predicted software stack and the cloud resource configuration.

12 . A method for real-time control of cloud computing resources allocated to an active cloud stack by a cloud resource control circuitry, comprising:
retrieving cloud stack configuration for the active cloud stack from a configuration repository; identifying a first collection of software components of a software stack of the active cloud stack as a upstream layer of the software stack based on the cloud stack configuration; identifying a second collection of software components of the software stack as a middle layer of the software stack based on the cloud stack configuration, wherein the middle layer consumes data generated by the upstream layer; and iteratively causing adjustment of computing resources allocated to the upstream layer and the middle layer of the software stack by:
com

Assignees

• Accenture Global Solutions Ltd

Inventors

• Das Jagaran
• Tanwar Nikhil S
• Zafar Abdul Rahman
• Elsherif Hossam E

Classifications

• H04L63/083

  using passwords (cryptographic mechanisms or cryptographic arrangements for entity authentication using a predetermined code H04L9/3226) · CPC title

• H04L67/1031

  Controlling of the operation of servers by a load balancer, e.g. adding or removing servers that serve requests · CPC title

• G06F9/5011 Primary

  the resources being hardware resources other than CPUs, Servers and Terminals · CPC title

• G06F2209/5019

  Workload prediction · CPC title

• G06F2209/508

  Monitor · CPC title