Cloud API Retrieval with Asynchronous Prefetch and Cache Fallback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for retrieving data objects in cloud computing environments are inefficient in handling numerous data retrieval operations, leading to high latencies due to the serial nature of synchronous calls, which burdens users with prolonged processing times.
Innovation Solution
Implementing a system that leverages asynchronous and synchronous application program interfaces (APIs) in tandem with a process flow, using an asynchronous controller to pre-fetch and cache data objects, while a synchronous controller handles specific data retrieval when necessary, thereby reducing latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If synchronous calls are used for data retrieval operations, then data can be retrieved reliably, but user latency increases due to the serial nature of the calls
Solution Approach 1:
The system performs preliminary data retrieval actions by pre-fetching data objects using asynchronous calls before they are actually needed. The asynchronous controller proactively retrieves data in advance and caches it, so when synchronous requests arrive, the data is already available, eliminating the serial waiting time while maintaining reliability.
Solution Approach 2:
An asynchronous controller is introduced as an intermediary between the data retrieval system and the synchronous requests. This intermediary handles the time-consuming data retrieval operations asynchronously, caching results so that subsequent synchronous requests can be fulfilled quickly without blocking, thus resolving the contradiction between reliability and latency.
2Loss of time
If asynchronous calls are used for data retrieval operations, then user latency is reduced, but data retrieval reliability may be compromised
Solution Approach 1:
The system performs preliminary data retrieval actions by pre-fetching data objects using asynchronous calls before they are actually needed. The asynchronous controller proactively retrieves data in advance and caches it, so when synchronous requests arrive, the data is already available, eliminating the serial waiting time while maintaining reliability.
Solution Approach 2:
An asynchronous controller is introduced as an intermediary between the data retrieval system and the synchronous requests. This intermediary handles the time-consuming data retrieval operations asynchronously, caching results so that subsequent synchronous requests can be fulfilled quickly without blocking, thus resolving the contradiction between reliability and latency.
3Measurement precision
If numerous data retrieval operations are handled serially, then data accuracy is maintained, but processing efficiency decreases
Solution Approach 1:
The system performs preliminary data retrieval actions by pre-fetching data objects using asynchronous calls before they are actually needed. The asynchronous controller proactively retrieves data in advance and caches it, so when synchronous requests arrive, the data is already available, eliminating the serial waiting time while maintaining reliability.
Solution Approach 2:
The system maintains continuous useful action by overlapping data retrieval operations with other processing tasks. While data is being fetched asynchronously in the background, the system can continue with other operations, and the retrieved data is automatically cached and ready for use, thereby maintaining data accuracy while dramatically improving processing efficiency.
Data Source
AI summary
Systems, methods, and devices are disclosed herein that provide management of usage of application program interfaces. A computing platform may be configurable to cause detecting a first action comprising an event at a user interface associated with an on-demand application hosted by the computing platform, and making, using a first application program interface (API), an asynchronous call for a data record in response to detecting the first action, the asynchronous call being made to a storage location of the computing platform. The computing platform may also be configurable to cause detecting a second action associated with the on-demand application, and determining if a synchronous call should be performed using a second API, the synchronous call being made in response to determining that a cache does not include the data record. The computing platform may also be configurable to cause generating a result responsive to the second action.


