GithubHelp home page GithubHelp logo

isabella232 / quic-trace Goto Github PK

View Code? Open in Web Editor NEW

This project forked from google/quic-trace

0.0 0.0 0.0 625 KB

Library and tools for transcribing QUIC connections.

License: Apache License 2.0

C++ 88.97% Shell 1.42% Starlark 9.61%

quic-trace's Introduction

QUIC trace utilities

This repository contains a format for recording a trace of QUIC connection, together with the tools for analyzing the resulting traces. This format of traces is currently partially supported by Chromium, and we hope for more implementations to adopt it in the future.

The primary focus of this format is debugging congestion-control related issues, but other transport aspects (e.g. flow control) and transport-related aspects of the application (e.g. annotations for the types of data carried by the streams) are also within the scope.

How to record this format

The traces are represented as a Protocol Buffer, which is completely described in lib/quic_trace.proto. Projects that use Bazel can embed this repository directly and use the provided Bazel rules.

OpenGL-based renderer

This repository contains two different tools to visualize the trace:

  1. A simple gnuplot-based trace renderer.
  2. A fully featured OpenGL-based renderer.

The OpenGL-based is the current preferred way of rendering traces. It supports Linux and macOS, though Windows support is coming soon. OpenGL 3.3 or later is required. Most of its dependencies are shipped using Bazel build files, but for Linux, system SDL2 is used.

OpenGL renderer screenshot

Build instructions

Building trace renderer on Linux:

  1. Install SDL2 with headers (libsdl2-dev on Debian and related systems).
  2. Install OpenGL headers (on Debian, libsdl2-dev pulls those in).
  3. Install Bazel from https://bazel.build
  4. Build the renderer by running bazel build -c opt //tools/render/...
  5. Run the renderer by running, for example, bazel-bin/tools/render/render_trace examples/example.qtr.

Building trace renderer on macOS:

  1. Install Bazel from https://bazel.build
  2. Build the renderer by running bazel build -c opt --apple_platform_type=macos --cpu=darwin_x86_64 //tools/render/...
  3. Run the renderer by running, for example, bazel-bin/tools/render/render_trace examples/example.qtr.

Note that macOS build requires a specific set of flags, because by default Bazel on macOS targets iOS.

Scalability

The OpenGL-based renderer was built with the goal of rendering large traces in mind, and our testing has shown that it performs quite well with traces as large as 120MiB worth of data transferred (that's approximately 90,000 packets sent), even when ran on a machine with high-resolution screen and an integrated GPU.

Legacy gnuplot-based renderer

This rendering script was originally built as a simple proof of concept for rendering the trace format, but still might be useful if other tools are not available. Instructions:

  1. Build the helper tool by running bazel build //tools/quic_trace_to_time_sequence_gnuplot.
  2. Generate the graph by running tools/time_sequence_gnuplot.sh trace_file trace.png.

quic-trace's People

Contributors

gfx avatar ianswett avatar janaiyengar avatar marten-seemann avatar ryantheoptimist avatar vasilvv avatar zyshi avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.