Evaluating Cloud-Native Storage: Performance Benchmarking for Uyuni on RKE2

I propose to build a reproducible benchmark framework for evaluating cloud-native storage options for Uyuni on RKE2. The problem is that different Kubernetes storage backends can behave very differently for the kinds of operations Uyuni depends on, especially PostgreSQL transactions, repository metadata handling, and package distribution. The project will address this by starting from a storage-only control baseline and then extending the evaluation into PostgreSQL benchmarking and Uyuni-specific workload benchmarking, while collecting synchronized observability data for each run. The same benchmark method will be reused across selected storage backends so that results remain comparable and can support a practical recommendation for which StorageClass Uyuni should support on RKE2. The main deliverables will be a reusable benchmark harness, benchmark results for the tested storage backends, PostgreSQL and Uyuni workload comparison data, observability-backed summaries, and a final report that explains the benchmark method, compares the backends, and recommends the most suitable StorageClass. The project will also produce the configuration artifacts and scripts needed to rerun the benchmark process and reproduce the tested setup.