Browse Prior Art Database

System and Method for Dynamic Resource Management Based on Risk Aware Metrics

IP.com Disclosure Number: IPCOM000248308D
Publication Date: 2016-Nov-15
Document File: 3 page(s) / 29K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for dynamic, intelligent, and self-learning resource management in the cloud based on risk in conjunction with location, bandwidth, latency, and cost.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 3

System and Method for Dynamic Resource Management Based on Risk Aware Metrics

Modern internet scale applications operate in virtual environments where the underlying resources required are constantly scaled, moved, and distributed across environments (local, dedicated cloud, public cloud) to maintain optimal performance and response times to the end users of the application. Managing these resources can be complicated as the location, bandwidth, latency, and various security policies all have to be taken into account, while maintaining low costs. When unexpected scenarios (e.g., failures, high-load, disasters, security hacks, etc.) happen, businesses are willing to absorb temporary risks to mitigate problems. Mitigation solutions include:


 Adding instances of the resource in the same operating environment


 Moving/adding instances in a similar environment in the same geographical region


 Moving/adding instances in a similar environment in a different geographical region


 Moving/adding instances to a different public cloud environment

Each of these solutions has limitations and restrictions that require consideration when creating rules. Businesses have multiple applications and each of those applications needs different policies.

Many cloud environments provide the operators the ability to define policies and rules to dynamically adjust resources based on several metrics, but it is not possible to associate quantitative risk to these policies to make these decisions smarter.

The novel contribution is a method for dynamic, intelligent, and self-learning resource management in the cloud based on risk in conjunction with location, bandwidth, latency, and cost.

The method works by allowing the user to provide a consistent way of defining an automatic scaling policy for managing resources across multiple applications. This includes:


 Adding instances of the recourse in the same operating environment


 Moving/adding instances in a similar environment in the same geographical region


 Moving/adding instances in a similar environment in a different geographical region


 Moving/adding instances in a different public cloud environment

Some applications are not suitable to run in certain geographies due to a higher risk of security attacks, government regulations around data residency, higher risk of failures, etc. However, in extreme cases, the business might be willing to absorb this risk and run an application, anyway. For this case, the user needs to provide a risk threshold for each of the scaling options, which instructs the auto-scaling engine to only perform this option as a last measure.

1


Page 02 of 3

For example:


1) Application X: Does not deal with any sensitive data and geographical restrictions are not much of a concern:

LOW - Adding additional instances of the recourse in the same operating environment
LOW - Moving/adding instances in a similar environment in the same geographical region
LOW - Moving/adding...