BENCHMARKING EMBEDDED KEY-VALUES STORES USING eBPF
Fuente:
Zenodo
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Pubblicazione: |
Zenodo
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866901926394724352 |
|---|---|
| author | Sahakyan H. |
| author_facet | Sahakyan H. |
| contents | <p><strong><span lang="EN-GB">Abstract</span></strong></p> <p><span lang="EN-GB">Embedded key-value stores run inside latency-critical paths, yet conventional benchmarks reveal only symptoms, not causes. We present ucsb-ebpf, a drop-in extension to the Unbranded Cloud Serving Benchmark that injects eBPF probes into every run. The tracer records per-thread syscall times and optional heap/slab events while adding under 1 % runtime overhead, so the same workload that measures throughput now pinpoints where each microsecond goes. Using seven standard mixes, we profile LevelDB, RocksDB, WiredTiger and LMDB on NVMe hardware. All code, data and analysis scripts are open sourced.</span></p> <p> </p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_15657638 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | BENCHMARKING EMBEDDED KEY-VALUES STORES USING eBPF Sahakyan H. eBPF; benchmarking; embedded key-value store; syscall tracing; memory profiling; performance analysis <p><strong><span lang="EN-GB">Abstract</span></strong></p> <p><span lang="EN-GB">Embedded key-value stores run inside latency-critical paths, yet conventional benchmarks reveal only symptoms, not causes. We present ucsb-ebpf, a drop-in extension to the Unbranded Cloud Serving Benchmark that injects eBPF probes into every run. The tracer records per-thread syscall times and optional heap/slab events while adding under 1 % runtime overhead, so the same workload that measures throughput now pinpoints where each microsecond goes. Using seven standard mixes, we profile LevelDB, RocksDB, WiredTiger and LMDB on NVMe hardware. All code, data and analysis scripts are open sourced.</span></p> <p> </p> |
| title | BENCHMARKING EMBEDDED KEY-VALUES STORES USING eBPF |
| topic | eBPF; benchmarking; embedded key-value store; syscall tracing; memory profiling; performance analysis |
| url | https://doi.org/10.5281/zenodo.15657638 |