Redisearch KNN benchmark

redisearch_knn_benchmark.go

package main

import (
	"context"
	"encoding/binary"
	"fmt"
	"math"
	"math/rand"
	"testing"
	"time"

	"github.com/RediSearch/redisearch-go/redisearch"
	"github.com/rueian/rueidis"
)

var INDEXNAME = "redisearch_benchmark"
var HOSTNAME = "127.0.0.1:6379"

func float32ToBytes(floats []float32) []byte {
	result := make([]byte, 0)
	for _, float := range floats {
		bytes := make([]byte, 4)
		bits := math.Float32bits(float)
		binary.LittleEndian.PutUint32(bytes, bits)
		result = append(result, bytes...)
	}
	return result
}

func BenchmarkByName(b *testing.B) {
	rand.Seed(time.Now().UnixNano())
	client, err := rueidis.NewClient(rueidis.ClientOption{InitAddress: []string{HOSTNAME}})
	if err != nil {
		panic(err)
	}

	// FT.CREATE redisearch_benchmark ON HASH PREFIX 1 movie: SCHEMA title TEXT year NUMERIC v VECTOR FLAT 6 TYPE FLOAT32 DIM 4 DISTANCE_METRIC L2
	ctx := context.Background()
	if err := client.Do(ctx, client.B().Arbitrary("FT.CREATE", INDEXNAME, "ON", "HASH", "PREFIX", "1", "movie:", "SCHEMA").
		Args("title", "TEXT").
		Args("year", "NUMERIC").
		Args("v", "VECTOR", "HNSW", "6", "TYPE", "FLOAT32", "DIM", "4", "DISTANCE_METRIC", "L2").Build()).Error(); err != nil {
		panic(err)
	}

	sampleNum := int(1e5)
	id := 0
	for id < sampleNum {
		now := time.Now().UnixNano()
		key := fmt.Sprintf("movie:matrix_%d", now)
		vec := float32ToBytes([]float32{rand.Float32(), rand.Float32(), rand.Float32(), rand.Float32()})

		title := fmt.Sprintf("Matrix%d", now)
		year := rand.Intn(100) + 1850

		if err := client.Do(ctx, client.B().Hset().Key(key).FieldValue().
			FieldValue("title", title).
			FieldValue("year", fmt.Sprintf("%d", year)).
			FieldValue("v", string(vec)).Build()).Error(); err != nil {
			panic(err)
		}

		id++
	}

	b.Run("rueidis", func(b *testing.B) {
		if err != nil {
			panic(err)
		}
		ctx := context.Background()
		b.ResetTimer()
		b.RunParallel(func(pb *testing.PB) {
			for pb.Next() {
				vec := float32ToBytes([]float32{rand.Float32(), rand.Float32(), rand.Float32(), rand.Float32()})
				_, err := client.Do(ctx, client.B().FtSearch().Index(INDEXNAME).
					Query("(@year:[1900 2000])=>[KNN 100 @v $B]").Sortby("__v_score").Asc().
					Limit().OffsetNum(0, 100).Params().Nargs(2).
					NameValue().NameValue("B", string(vec)).Dialect(2).Build()).ToArray()
				if err != nil {
					b.Fatal(err)
				}
			}
		})
		b.StopTimer()
	})

	b.Run("redigo", func(b *testing.B) {
		c := redisearch.NewClient(HOSTNAME, INDEXNAME)
		b.ResetTimer()
		b.RunParallel(func(pb *testing.PB) {
			for pb.Next() {
				vec := float32ToBytes([]float32{rand.Float32(), rand.Float32(), rand.Float32(), rand.Float32()})
				q := redisearch.NewQuery("(@year:[1900 2000])=>[KNN 100 @v $B]").SetSortBy("__v_score", true).
					Limit(0, 100).SetParams(map[string]interface{}{"B": vec}).SetDialect(2)
				_, _, err := c.Search(q)
				if err != nil {
					b.Fatal(err)
				}
			}
		})
		b.StopTimer()
	})

	client.Close()
	fmt.Println("DONE")
}

KNN Benchmark for official redisearch go client and rueidis client.

redisearch config.

# THREADED I/O
io-threads 4
io-threads-do-reads yes

run command

docker run -p 6379:6379 -d --name test-redisearch \
    -v redis.conf:/etc/redis/redis.conf \
    redislabs/redisearch:2.4.12

test result:

Seems no significant improvement for FT.SEARCH when using the auto pipeline.