← zurück zur Übersicht

2026-06-15T22-08-08Z_pgvector-S0-hybrid-r1

pgvectorStufe S0 · 2.00 GBhybridStatus: ok

Metriken

Throughput
112.7 QPS
Latenz ⌀
8.84 ms
Latenz p50
4.49 ms
Latenz p95
9.95 ms
Latenz p99
92.10 ms
Recall@1
11.60%
Recall@10
32.68%
Recall@100
38.37%
Precision@10
32.68%
NDCG@10
43.59%

Rohdaten

Konfiguration

Config-Name
pgvector-S0-hybrid
DB-Version
Image
Index-Typ
hnsw
Index-Params
{"m":16,"ef_construction":128,"ef_search":64}
Dataset
S0 · 550.000 Vektoren · dim 1024 · Variante A · 2.00 GB
Queries
2.000 (Concurrency 1)

Lauf

Gestartet
16.06.2026, 00:08:08
Beendet
16.06.2026, 00:08:32
Dauer
7224s
Index-Bauzeit
666.04s
Index-Größe
4249.4 MB
CPU ⌀ / Peak
0.40 / 0.42 cores
RAM ⌀ / Peak
6850 / 6922 MB
Disk read (Paging)
1 MB
Disk write
0 MB
Pod-RAM-Limit
Default
Warmup-Queries
1.000
K8s
v1.31.4+k3s1
Nodes
4

Notizen

measured: in-cluster
decoupled: true
ingested_at: 2026-06-15T22:04:31Z
ingest_config: pgvector-S0-latency
insert_time_s: 224.37
n_vectors_actual: 550000
has_metadata: true
mem_limit_gb: null
pre_run_reset: skipped (decoupled, warmup-discard)
n_queries_executed: 1000
n_warmup: 1000
gt_file: ground_truth_hybrid_alpha_50_ids.npy
mode: measure-only
repeat_group: 2026-06-15T22-08-08Z_pgvector-S0-hybrid
repeat_index: 1
repeat_total: 6