2026-06-19T06-36-21Z_pgvector-S1-tune-hybrid-a10-r5
pgvectorStufe S1 · 5.00 GBhybridStatus: ok
Metriken
Throughput
8.4 QPS
Latenz ⌀
118.42 ms
Latenz p50
102.28 ms
Latenz p95
183.39 ms
Latenz p99
293.92 ms
Recall@1
2.40%
Recall@10
2.88%
Recall@100
3.57%
Precision@10
2.88%
NDCG@10
3.37%
Rohdaten
summary.json ↗Vollständige Run-Metrikenhttps://raw.githubusercontent.com/unrealshape/bachelor-db-benchmark/main/results/2026-06-19T06-36-21Z_pgvector-S1-tune-hybrid-a10-r5/summary.jsonconfig.json ↗Eingesetzte Konfigurationhttps://raw.githubusercontent.com/unrealshape/bachelor-db-benchmark/main/results/2026-06-19T06-36-21Z_pgvector-S1-tune-hybrid-a10-r5/config.jsonDatensatz · 5.00 GB ↗Generator + Bauanleitung im RepoRun-Ordner auf GitHub ↗Beide Rohdateien im Verzeichnis
Konfiguration
- Config-Name
- pgvector-S1-tune-hybrid-a10
- DB-Version
- —
- Image
- —
- Index-Typ
- ivfflat
- Index-Params
- {"lists":1200,"probes":10}
- Dataset
- S1 · 1.350.000 Vektoren · dim 1024 · Variante A · 5.00 GB
- Queries
- 1.000 (Concurrency 1)
Lauf
- Gestartet
- 19.06.2026, 08:36:21
- Beendet
- 19.06.2026, 08:40:23
- Dauer
- 7442s
- Index-Bauzeit
- 160.1s
- Index-Größe
- 10556.3 MB
- CPU ⌀ / Peak
- 0.44 / 1.31 cores
- RAM ⌀ / Peak
- 7450 / 7515 MB
- Disk read (Paging)
- 32895 MB
- Disk write
- 72 MB
- Pod-RAM-Limit
- 8 GiB
- Warmup-Queries
- 1.000
- K8s
- v1.31.4+k3s1
- Nodes
- 4
Notizen
measured: in-cluster decoupled: true ingested_at: 2026-06-19T06:19:44Z ingest_config: pgvector-S1-ivfflat insert_time_s: 541.61 n_vectors_actual: 1350000 has_metadata: true mem_limit_gb: 8 pre_run_reset: skipped (decoupled, warmup-discard) n_queries_executed: 1000 n_warmup: 1000 gt_file: ground_truth_hybrid_alpha_10_ids.npy mode: measure-only repeat_group: 2026-06-19T06-19-44Z_pgvector-S1-tune-hybrid-a10 repeat_index: 5 repeat_total: 6