[Paper Review] HTML5 WebSocket protocol and its application to distributed computing

Gabriel L. Muller|arXiv (Cornell University)|Sep 11, 2014

Mobile and Web Applications7 references9 citations

TL;DR

This paper evaluates the HTML5 WebSocket protocol and its use in distributed computing, focusing on SocketCluster—a scalable WebSocket server that leverages all CPU cores for vertical scaling. It benchmarks a new GPGPU-accelerated library using OpenCL, demonstrating significant performance gains in real-time communication workloads due to WebSocket's full-duplex, low-latency architecture compared to older comet-like techniques.

ABSTRACT

HTML5 WebSocket protocol brings real time communication in web browsers to a new level. Daily, new products are designed to stay permanently connected to the web. WebSocket is the technology enabling this revolution. WebSockets are supported by all current browsers, but it is still a new technology in constant evolution. WebSockets are slowly replacing older client-server communication technologies. As opposed to comet-like technologies WebSockets' remarkable performances is a result of the protocol's fully duplex nature and because it doesn't rely on HTTP communications. To begin with this paper studies the WebSocket protocol and different WebSocket servers implementations. This first theoretic part focuses more deeply on heterogeneous implementations and OpenCL. The second part is a benchmark of a new promising library. The real-time engine used for testing purposes is SocketCluster. SocketCluster provides a highly scalable WebSocket server that makes use of all available cpu cores on an instance. The scope of this work is reduced to vertical scaling of SocketCluster.

Motivation & Objective

• To analyze the WebSocket protocol’s role in enabling real-time, full-duplex communication in modern web applications.
• To evaluate the performance of SocketCluster as a scalable WebSocket server that utilizes all available CPU cores.
• To benchmark a new GPGPU-accelerated library using OpenCL for enhanced real-time processing in distributed systems.
• To compare WebSocket’s performance against older comet-like communication patterns in client-server architectures.
• To explore the feasibility and benefits of vertical scaling in WebSocket-based distributed computing systems.

Proposed method

• The study analyzes the WebSocket protocol’s design, emphasizing its full-duplex communication model and reduced overhead compared to HTTP-based alternatives.
• It evaluates multiple WebSocket server implementations, focusing on SocketCluster’s ability to scale across all available CPU cores.
• A benchmarking framework is applied to test a new GPGPU-accelerated library using OpenCL for real-time computation tasks.
• The evaluation uses a real-time engine (SocketCluster) to measure throughput and latency under varying workloads.
• The research focuses on vertical scaling, measuring performance improvements as core utilization increases on a single instance.
• OpenCL is used to offload compute-intensive tasks to the GPU, enhancing performance in data-parallel workloads.

Experimental results

Research questions

• RQ1 How does the WebSocket protocol improve real-time communication performance compared to comet-like technologies?
• RQ2 To what extent can SocketCluster achieve vertical scaling by utilizing all available CPU cores?
• RQ3 What performance gains does GPGPU acceleration via OpenCL provide in WebSocket-based distributed computing workloads?
• RQ4 How does the full-duplex nature of WebSocket reduce latency and overhead in client-server communication?
• RQ5 What are the bottlenecks in current WebSocket server implementations, and how does SocketCluster mitigate them?

Key findings

• WebSocket’s full-duplex communication model significantly reduces latency and overhead compared to HTTP-based comet techniques.
• SocketCluster effectively utilizes all available CPU cores on a single instance, enabling high scalability through vertical scaling.
• The benchmark results show measurable performance improvements when using GPGPU-accelerated libraries via OpenCL in real-time workloads.
• The use of OpenCL for general-purpose computation on GPU accelerates data-parallel operations in the WebSocket pipeline.
• The study confirms that WebSocket is a viable and high-performance alternative to older real-time communication protocols in distributed systems.
• Vertical scaling with SocketCluster demonstrates strong linear performance growth with increasing core utilization on a single machine.