Hybrid Pull Push Streaming Reduces Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing pull-based methodology for delivering live content over the Internet and packet-switched networks introduces significant delays due to the need to wait for entire stream segments to be uploaded and stored, leading to cache misses and round-trip times that cause playback to commence several seconds after a corresponding broadcast feed.
Innovation Solution
The hybrid pull and push based streaming method employs a simplified CDN architecture that prepopulates caches with stream segment data as it is uploaded, allowing immediate serving of requested segments without cache misses, and transitions from pull-based to push-based distribution based on demand, reducing latency and delay.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pull-based methodology is used to deliver live content, then client devices can request stream segments from CDN edge PoPs, but significant delays occur due to cache misses and round-trip times to origin PoP and storage
Solution Approach 1:
The system performs preliminary actions by pushing stream segment data to edge PoP caches before client devices request them. The encoder uploads segments to the origin PoP, which then proactively pushes them to edge PoPs in advance, so when clients request segments, they are already available in the cache, eliminating the need for round-trip requests to the origin PoP and reducing latency significantly.
Solution Approach 2:
The origin PoP acts as an intermediary between the encoder and edge PoPs. Instead of clients directly requesting segments from storage or origin PoP (causing long round-trip delays), the origin PoP receives segments from the encoder and pushes them to edge PoPs, serving as a mediator that enables faster local delivery to clients while maintaining reliable stream delivery through the CDN infrastructure.
2Reliability
If entire stream segments are uploaded and stored before delivery, then complete segment data is available for redistribution, but playback commences several seconds after broadcast feed due to encoding and upload time
Solution Approach 1:
The system performs preliminary action by pushing segment data to edge PoP caches as soon as the origin PoP receives it from the encoder, before client requests arrive. This proactive pushing ensures that when clients request segments, the data is already prepared and available locally, eliminating waiting time while ensuring complete segment data is delivered through the push mechanism.
Solution Approach 2:
The system maintains continuity of useful action by establishing a continuous push workflow where the origin PoP continuously pushes segments to edge PoPs as they are received from the encoder. This continuous operation eliminates idle waiting periods and ensures that segment delivery progresses without interruption, reducing the overall delay between encoding and client playback.
3Loss of time
If push-based distribution is used for high demand streams, then latency is reduced by prepopulating caches, but bandwidth consumption increases for distributing segments proactively
Solution Approach 1:
The system applies dynamics by making the distribution strategy adaptive rather than static. The origin PoP monitors demand metrics (such as request rates or viewer counts) and dynamically switches between push-based and pull-based distribution for different streams or different segments. For high-demand streams, push-based distribution prepopulates caches to reduce latency; for low-demand streams, pull-based distribution conserves bandwidth by only transferring segments when requested. This dynamic adaptation resolves the contradiction between latency reduction and bandwidth consumption.
Data Source
AI summary
Hybrid pull and push based streaming selectively performs a pull-based distribution of a stream to a first point-of-presence (“PoP”) of a distributed platform having low demand for the stream, and a push-based distribution of the stream to a second PoP of the distributed platform having high demand for the stream. The push-based distribution may be used to prepopulate the second PoP cache with the live stream data as the live stream data is uploaded from an encoder to a source PoP of the distributed platform, and before that live stream data is requested by the second PoP. In doing so, requests for the live stream data received at the second PoP may result in cache hits with the requested live stream data being immediately served from the second PoP cache without having to retrieve the live stream data from outside the second PoP.