Ephemeral Object Push to Edge Server Cache
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional Content Delivery Networks (CDNs) are inefficient for distributing ephemeral objects that are read once or very few times, as they rely on a pull-based model that introduces additional latency and are not optimized for such scenarios.
Innovation Solution
A computer-implemented method that evaluates object push criteria to proactively push ephemeral objects to edge server caches based on specified viewing periods, number of views, or time, allowing for speculative population and serving requests with reduced latency, using object push criteria like traffic patterns, historical use, social graphs, and geolocation information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a pull-based model is used to serve objects from edge server caches, then service efficiency is improved for objects that are read many times, but latency increases for objects that are read once or very few times
Solution Approach 1:
The system performs preliminary actions by pushing ephemeral objects to edge server caches before client requests arrive. The origin server proactively distributes objects to multiple edge servers based on push criteria, so when a client requests an object, it can be served immediately from the nearest edge server without waiting for a pull operation, thereby reducing latency for infrequently accessed objects
Solution Approach 2:
The system inverts the traditional pull-based CDN model by implementing a push-based approach for ephemeral objects. Instead of waiting for client requests to pull objects from the origin server, the origin server actively pushes objects to edge servers in advance, reversing the direction of data flow optimization to benefit short-lived, infrequently accessed content
2Speed
If objects are cached in edge servers using traditional CDN models, then delivery speed is improved for frequently accessed objects, but cache space is wasted on objects that are read once or very few times
Solution Approach 1:
The system performs preliminary distribution of ephemeral objects to multiple edge servers before client requests arrive. By pushing objects to edge servers in advance based on push criteria (geolocation, traffic patterns, historical use), the system ensures fast delivery when requests occur while minimizing wasted cache space since objects are only pushed to relevant edge servers rather than universally caching all objects
Solution Approach 2:
The system applies local quality by pushing different ephemeral objects to different edge servers based on location-specific push criteria. Each edge server receives a customized set of objects relevant to its geographic region and expected traffic patterns, optimizing cache utilization locally rather than uniformly distributing all objects to all edge servers
3Reliability
If a pull-based model is used, then cache hit performance is optimized for large objects read many times, but additional latency is introduced for ephemeral objects
Solution Approach 1:
The system performs preliminary population of edge server caches with ephemeral objects before client requests arrive. The origin server pushes objects to edge servers based on push criteria, so when a client requests an ephemeral object, it is already cached at the nearest edge server and can be served immediately, eliminating the latency associated with pull-based fetching while maintaining reliable cache hit performance
Data Source
AI summary
A computer implemented method includes receiving an object scheduled for automatic deletion after a specified viewing period, a specified number of views or a specified period of time. Object push criteria are evaluated. The object is pushed to an edge server cache in response to evaluating. The object is served in response to a request for the object.