A "Cloud Sourcing" System that Recommends the Pptimal Resource Configuration Based on a Cloud Workload's Attributes
Publication Date: 2014-Apr-15
The IP.com Prior Art Database
Disclosed is a system that determines what types of physical resources and software products to provision on a cloud, based on indirect attributes identified by the customer. The solution captures various attributes required by the end user for a specific business, identifies the technical requirements, and then provisions the overall cloud infrastructure leveraging a decision support system.
Page 01 of 5
A "Cloud Sourcing" System that Recommends the Pptimal Resource Configuration
Based on a Cloud Workload's Attributes
Cloud sourcing system provides no direct integration of of the infrastructure services to managed service providers. Current solutions have fixed images of services with built in agents or applications, but do not have direct integration with services providers. The infrastructure must be compiled based on these fixed components and are not custom built based on client required attributes.
Current cloud deployments are Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS), and Business Process as a Service (BPaaS). End user requirements may vary based on the application performance, capacity, availability, security, monitoring, service levels and price. A system is needed that provisions cloud based infrastructure components (i.e. products, services) from various service providers based on attributes chosen by the end user.
The novel contribution is a system and method that determines what types of physical resources and software products to provision on a cloud, based on indirect attributes identified by the customer. The solution captures various attributes required by the end user for a specific business, identifies the technical requirements, and then provisions the overall cloud infrastructure leveraging a decision support system. The decision support system is built based on product features, services features, price, service levels etc. offered by various vendors.
The solution comprises two methods: (1) comparison and analysis (2) determine best-fit software products to meet the customer's needs.
The first method compares a cloud customer's business attributes to existing customers' attributes and analyzes an optimal resource configuration based on existing customers' changes over time. This method uses a customer's attribute selection (see top part of figure 1 below) and to determine what types of hardware resources to allocate when provisioning on the cloud. Instead of specifically asking for resources (e.g., three virtual machines (VMs) with 8GB of Random Access Memory (RAM), four Central Processing Units (CPUs), and 200GB of disk, etc.) a customer only identifies a set of business attributes (determined by the cloud provider). The system then finds other workloads having similar attributes and analyzes those workloads to find out which hardware resources is an optimal choice. If 10 other cloud customers have workloads with the same or similar attributes, the method then analyzes what hardware changes were made over time (e.g., adding more CPU, reducing memory, etc.) and find patterns that indicate stability (e.g., few or no changes in the last X months). It then averages the result set of these ideal customers with matching attributes. Finally, the system can recommend a similar device/hardware configuration for the new customer request.
Figure 1: Example of Select...