Master Thesis: Benchmarking Cloud-Native Networking Solutions

Stack

Responsibilities

• ▹Investigate the networking stacks used in production: Linux kernel networking, SR-IOV and DPDK
• ▹Investigate I/O mechanisms and their trade-offs: sendmsg/recvmsg, sendmmsg/recvmmsg and io_uring
• ▹Investigate Kubernetes networking topologies: containers within the same pod, pods on the same worker node and pods across worker nodes
• ▹Investigate CNI plugins and the performance impact of network policy enforcement
• ▹Measure throughput (pps/bps), latency, CPU usage and system call overhead
• ▹Analyse the results and evaluate the suitability of each solution for different workload profiles
• ▹Present the results at an Ericsson technology forum

Requirements

• ▹Currently pursuing a Master's degree in Computer Science, Computer Engineering, Electrical Engineering or a related field
• ▹Interest in high-performance networking and cloud-native technologies
• ▹Basic programming experience in C or C++
• ▹Familiarity with shell scripting
• ▹Experience using Linux through studies or personal projects
• ▹Basic understanding of socket programming

Nice to have

• ▹Introductory knowledge of Kubernetes or containerization
• ▹Familiarity with DPDK or SR-IOV
• ▹Experience with networking projects through coursework or personal projects

Soft skills

About the company

Ericsson is a market leader in mobile network solutions, building the next generation of cloud-native Packet Core.