Cloud Agnostic Task Scheduler Batching API Requests
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for servicing API requests in cloud environments are inefficient and not scalable, leading to slow processing and potential timeouts, especially when handling a large number of services.
Innovation Solution
A cloud agnostic task scheduler that batches and schedules requests by identifying combinable tasks, converting them into combined tasks, and sending them for execution, while maintaining concurrency and preserving the order of operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If API requests are serviced sequentially one at a time, then the ordering of commands is preserved, but the processing speed is slow and inefficiency occurs
Solution Approach 1:
The patent implements a pre-execution validation phase where the scheduler analyzes the batch of API requests before actual execution. It determines the execution order, validates dependencies, and checks for conflicts among all requests in the batch beforehand. This preliminary analysis enables the system to execute multiple requests in parallel while guaranteeing correct ordering through pre-computed execution plans, thus resolving the contradiction between maintaining command ordering and improving processing speed.
2Device complexity
If conventional sequential techniques are used, then implementation simplicity is maintained, but scalability is limited and cannot accommodate a large number of services
Solution Approach 1:
The patent creates a universal batch scheduling framework that can handle diverse API requests from multiple cloud services through a unified approach. The scheduler uses a standardized interface to process different types of requests (create, delete, update) across various services, making the system scalable to accommodate large numbers of services without requiring service-specific implementation complexity. This universal framework resolves the contradiction by enabling scalability while keeping the core implementation relatively simple through abstraction.
3Use of energy by moving object
If API requests are processed one at a time, then resource consumption per request is controlled, but overall processing time increases and timeouts occur
Solution Approach 1:
The patent merges multiple individual API requests into a single batch that is processed together. By combining requests that can be executed in parallel into one batch operation, the system reduces the total number of sequential processing cycles needed. The batched approach maintains controlled resource consumption through unified resource allocation while dramatically reducing overall processing time by eliminating the sequential overhead and enabling parallel execution, thus resolving the contradiction between resource control and time efficiency.
Data Source
AI summary
Cloud agnostic task scheduling is disclosed. In an embodiment, a method includes receiving a configuration request for configuring a virtual service in a cloud environment. In response to a pre-specified event, one or more objects that are operated on by the configuration request is identified. The method determines whether the request can be processed based at least in part on a list of pending tasks. In response to a determination that the request can be processed, the method derives one or more end states for the one or more objects that are identified. The method converts the request to a set of one or more tasks operating on the one or more objects, identifies tasks in the list of pending tasks that are combinable, combines the identified tasks into a combined task that would result in the one or more end states, and sends the combined task to be executed.