applications open

Open UG Labs Micro-Fellowship: HTTP Performance Benchmarking for the urunc Evaluation Suite

A two-week mentored open-source contribution program where one selected student will extend the urunc Evaluation Suite with support for HTTP network performance benchmarking.

Apply now Deadline 1 October 2026 at 02:00

About the program

Open UG Labs Micro-Fellowship is a two-week intensive contribution program organized by Open UG Labs. The selected participant will contribute to the urunc Evaluation Suite, an open-source project used to evaluate and compare the performance of container and sandboxed runtimes. Project repository: https://github.com/urunc-dev/evaluation_suite The focus of this program will be extending the evaluation suite with support for HTTP application-layer network performance benchmarking. This work continues from previous development and research carried out during the 2026 LFX Mentorship on urunc. The selected contributor will first become familiar with the existing evaluation suite and its networking benchmarks. They will then work on designing and implementing support for running HTTP benchmarks, collecting useful performance measurements, validating the implementation, documenting the new functionality, and preparing the contribution for upstream review. The participant does not need previous experience with urunc, container runtimes, or HTTP benchmarking. The program is intended to provide an opportunity to learn these concepts while making a practical open-source contribution. Basic programming knowledge, familiarity with Linux, and willingness to learn are more important than previous experience with the specific technologies involved. To prepare for the program, applicants are encouraged to become comfortable running containers using containerd and nerdctl, particularly creating, starting, inspecting, and networking containers from the command line. Applicants are also encouraged to read about HTTP performance benchmarking and common metrics such as requests per second, throughput, response latency, latency percentiles such as p50, p95 and p99, error rates, timeouts, and concurrent connections. It would also be helpful to experiment with one or more commonly used HTTP benchmarking tools. Some useful resources include wrk2 at https://github.com/giltene/wrk2, Fortio at https://fortio.org/, and hey at https://github.com/rakyll/hey. You are not expected to master these tools before applying. These resources are provided to help interested applicants understand the type of work involved and prepare themselves before the program begins. Throughout the two weeks, the participant will receive mentorship and technical guidance. They will be expected to communicate their progress, ask questions when they encounter challenges, respond to technical feedback, and work toward completing the agreed deliverables within the program period. Successful completion will be based on the quality and completion of the agreed implementation, testing, documentation, and submission of the contribution for upstream review. The contribution does not need to be merged during the two-week program, since upstream review and merge timelines may depend on external maintainers.