Crusader: A Modern Tool for Deep Network Diagnostics
For network professionals and power users, getting a single bandwidth number from a typical speed test is rarely enough. When troubleshooting congestion, bufferbloat, or poor Quality of Service (QoS), you need a specific tool. One that can run complex, multi-stream tests and correlate throughput with latency. This is where specialized tools like FLENT and the newer Crusader Network Tester come in.
Crusader is a modern, cross-platform application designed to measure network throughput, latency, and packet loss specifically in the presence of simultaneous upload and download traffic.
How Crusader Compares to FLENT
Crusader and FLENT share a core philosophy: they both focus on analyzing network performance under congestion to reveal real-world bottlenecks. Both require a client and server setup to provide accurate, two-way measurements.
The main difference lies in their underlying technology and accessibility. FLENT is a Python-based wrapper that orchestrates established tools like netperf, excelling in its use of standard, complex test suites like Realtime Response Under Load (RRUL) for rigorous bufferbloat detection.
Crusader is built in Rust, which often allows for self-contained, high-performance binaries with minimal runtime dependencies. This difference in architecture allows Crusader to offer pre-built graphical user interfaces (GUIs) for Windows, Mac, and Android, making the initial setup and visualization process significantly more user-friendly compared to FLENT’s command-line interface and dependency management.
Motivation and Major Use Cases
The primary motivation behind Crusader is to provide a comprehensive, repeatable test. A test that clearly visualizes the correlation between data transfer and network responsiveness. High-bandwidth transfers can cause latency to spike dramatically due to bufferbloat, but simple speed tests often use streams that are too optimized to reveal this problem.
Major use cases for Crusader include:
- Diagnosing Bufferbloat: The simultaneous display of high throughput and corresponding latency spikes allows users to visually quantify the effect of congestion on real-time traffic.
- Comparative Network Testing: Easily test and compare the performance profiles of different network hardware. For example: routers, modems, or VPN configurations.
- End-to-End Performance Monitoring: The built-in Monitor tab offers a continuous latency check, providing an ongoing health check for your connection’s background responsiveness.
Installation and Setup
One of Crusader’s biggest draws is its straightforward installation. You need to run the client on one computer and the server on another (or a remote server).
- Pre-built Binaries: For Windows, macOS, and Android, simply navigate to the Releases page on the GitHub repository and download the appropriate pre-built executable. The GUI version contains both the client and server functionality.
- Linux: While a command-line client binary is available, you must build the GUI from the source. Alternatively, a pre-built Docker container is provided for easily running the Crusader server.
Example Usage and Test Profile
A Crusader test requires you to first start the Server on one machine and then use the Client GUI on the second machine, pointing it at the server’s IP address.
By default, the client runs a profile designed to stress the connection in three distinct 10-second phases, each separated by a period of idle time:
- Download-only Burst: Measures throughput and latency while only downloading data.
- Upload-only Burst: Measures throughput and latency while only uploading data.
- Bi-directional Burst: Measures throughput, latency, and packet loss with simultaneous upload and download traffic.
Results Tab
Upon completion, the Result tab generates three synchronized plots: Throughput, Latency, and Packet Loss. The Latency plot is particularly useful, showing uni-directional times (Server-to-Client, Client-to-Server) and the Round-Trip Time, allowing for precise identification of where the congestion pressure originates.
I tested this by running the server on my Windows machine and the client on my MacOS.
The server side is pretty bare bones, with just some log lines showing major events such as the server launch and the start and completion of client connections:
On the client side you can see graphs the download, upload, and bidirectional testing with the corresponding latency and packet loss graphs:
An additional feature of Crusade is that you can run a test on the browser by starting the server on “Remote” mode as follows:
On the client machine, you can access the test page that looks like this:
And once the test is finished, you can see the results on your browser:
The above test was run between the same MacOS (client) and Windows (server) machines. You can see that while in the previous test the download/upload were close to 1Gbps, on this version it’s around 120Mbps. Reason being, this test is executed on the browser, and not with the native Crusade client on the MacOS client.
Conclusion
Crusader Network Tester provides a highly accessible, modern, and visually rich platform for deep network diagnostics. By offering pre-built GUIs and focusing its test profile on the critical areas of latency under load and bi-directional performance, it serves as an excellent, user-friendly alternative to more complex, command-line-driven tools. For anyone serious about understanding the true quality and stability of their network, especially when battling the common foe of bufferbloat, Crusader is a powerful new weapon in the diagnostic arsenal.
