Serverless Deployment Artifact Grouping for FaaS CaaS Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cloud computing deployments of serverless applications face challenges in optimizing resource utilization and handling workload fluctuations across FaaS and CaaS platforms, with existing solutions either being inflexible for complex applications or inefficient in managing runtime costs.
Innovation Solution
An apparatus and method that dynamically determine a modified cloud computing deployment by grouping application functions into deployment artifacts associated with FaaS and CaaS platforms, optimizing their association based on application models and requirements, and using equation solvers to ensure performance and cost constraints are met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If FaaS platforms are used for instantaneous detection and handling of workload fluctuations, then responsiveness to workload changes is improved, but implementation complexity for complex applications with many interrelated functions increases
Solution Approach 1:
The patent segments complex applications into multiple deployment artifacts, each containing specific application functions. This segmentation allows FaaS platforms to handle individual functions independently, improving responsiveness to workload changes while managing implementation complexity through modular organization of application components.
Solution Approach 2:
The patent implements dynamic configuration of deployment artifacts, where the system can automatically adjust which functions are grouped together and deployed to which platform based on runtime conditions. This dynamic approach enables the system to optimize for responsiveness when needed while maintaining manageability through automated reconfiguration.
2Adaptability or versatility
If CaaS orchestration platforms are configured to automatically handle workload fluctuations, then flexibility in software deployment is improved, but the process of detecting workload changes and instantiating new containers cannot be performed instantaneous, resulting in runtime costs
Solution Approach 1:
The patent divides applications into deployment artifacts that can be selectively deployed to FaaS or CaaS platforms. This segmentation enables time-critical functions to be handled by FaaS for instantaneous response while less time-sensitive functions remain on CaaS, thus reducing overall runtime costs while maintaining flexibility.
Solution Approach 2:
The patent introduces an intermediary configuration system that manages the mapping between deployment artifacts and platform types. This intermediary layer provides flexibility in deployment decisions while enabling faster response times by routing appropriate functions to FaaS platforms that can instantiate and scale more quickly.
3Productivity
If application functions are grouped into deployment artifacts associated with specific platform types, then deployment optimization is improved, but the complexity of managing associations across multiple platform types increases
Solution Approach 1:
The patent creates a universal configuration framework that can manage deployment artifacts across multiple platform types (FaaS and CaaS) through a common interface. This universal approach enables deployment optimization for each platform while managing complexity through standardized configuration mechanisms that work across different platform types.
Data Source
AI summary
A technique for dynamically determining a modification of an initial cloud computing deployment (CCD) of a serverless application with multiple application functions is described. The multiple application functions in the initial CCD are grouped into one or more deployment artifacts each comprising at least one application function, wherein each deployment artifact is associated with a dedicated cloud computing platform type selected from FaaS and CaaS. An apparatus of the present disclosure is configured to obtain at least one requirement for the serverless application or its deployment, and to obtain an application model of the serverless application. The apparatus is further configured to determine, based on the application model and the at least one requirement, a modified CCD of the serverless application, wherein the modified CCD comprises at least one of a modified grouping of the multiple application functions into one or more deployment artifacts and a modified association of a particular deployment artifact with a dedicated cloud computing platform type selected from at least FaaS and CaaS.