elasticsearch向量数据库服务搭建
elasticsearch向量数据库服务搭建
文章目录
一、es安装步骤
1、创建网络
docker network create es-net
2、拉取镜像
docker pull elasticsearch:8.15.0
docker pull docker.elastic.co/elasticsearch/elasticsearch:8.15.0
3、创建挂载目录
mkdir -p /home/data/elasticsearch/config
mkdir -p /home/data/elasticsearch/data
mkdir -p /home/data/elasticsearch/logs
mkdir -p /home/data/elasticsearch/plugins
4、赋予权限
chmod -R 777 /home/data/elasticsearch
5、创建配置（可选项）
vi /home/data/elasticsearch/config/elasticsearch.yml
# 可访问IP
cluster.name: "my-es"
network.host: 127.0.0.1
# 集群的节点列表（ElasticSearch8新配置）
# discovery.seed_hosts: ["es-ip:9300"]
# 跨域开启密码
http.cors.enabled: true
http.cors.allow-origin: "*"
# http.cors.allow-headers: Authorization
xpack.security.enabled: false
6、创建脚本
vim startEs.sh
sudo docker run -d \
--restart=always \
--name elasticsearch \
--network es-net \
-p 9200:9200 \
-p 9300:9300 \
--privileged \
-v /home/data/elasticsearch/data:/usr/share/elasticsearch/data \
-v /home/data/elasticsearch/plugins:/usr/share/elasticsearch/plugins \
-e "discovery.type=single-node" \
-e "ES_JAVA_OPTS=-Xms512m -Xmx512m" \
elasticsearch:8.15.0
docker exec -it elasticsearch /bin/bash
# 关闭 密码安全验证
echo 'xpack.security.enabled: false' >> elasticsearch.yml
# 重启容器
sudo docker restart elasticsearch
7、测试Elasticsearch是否安装成功
http://192.168.198.197:9200
8、容器操作
# 进入容器
docker exec -it elasticsearch /bin/bash
sudo docker exec -it elasticsearch /bin/bash
# 重启容器
docker restart elasticsearch
二、es可视化kibana安装步骤
1、拉取镜像
docker pull kibana:8.10.2
2、创建配置（可选项）
vi /home/data/kibana/config/kibana.yml
# 主机地址，可以是ip,主机名
server.host: 0.0.0.0
# 提供服务的端口，监听端口
server.port: 5601
# 该 kibana 服务的名称，默认 your-hostname
server.name: "bolei-kibana"
server.shutdownTimeout: "5s"
#####----------elasticsearch相关----------#####
# kibana访问es服务器的URL,就可以有多个，以逗号","隔开
elasticsearch.hosts: [ "http://192.168.198.197:9200" ]
monitoring.ui.container.elasticsearch.enabled: true
3、启动镜像
sudo docker run -d \
--name kibana2 \
-e ELASTICSEARCH_HOSTS=http://elasticsearch:9200 \
--network=es-net \
-p 5601:5601 \
kibana:8.6.0
4、容器炒作
# 进入容器
docker exec -it kibana2 /bin/bash
sudo docker exec -it kibana2 /bin/bash
# 重启容器
docker restart kibana1
5、访问地址
http://192.168.198.197:5601
三、安装elasticsearch-head
直接使用浏览器插件Multi ElasticSearch Head
使用教程：https://blog.csdn.net/qq_50854662/article/details/135967448
四、数据测试，kibana脚本，操作es
dev tool控制台处理
# 创建索引
PUT goods_index
# 查询索引
GET goods_index
# 添加映射
PUT goods_index/_mapping
{
"properties":{
"age":{
"type":"integer"
},
"name":{
"type":"keyword"
},
"desc":{
"type":"text"
}
}
}
# 创建索引并且添加映射
PUT goods_index1
{
"mappings": {
"properties": {
"age":{
"type":"integer"
},
"name":{
"type":"keyword"
},
"desc":{
"type":"text"
}
}
}
}
# 添加文档，指定id
PUT goods_index/_doc/1
{
"name":"张三",
"age":34,
"desc":"北京人深圳来"
}
# 添加文档，不指定id
POST goods_index/_doc
{
"name":"李四",
"age":25,
"desc":"四川妹子真辣"
}
# 修改文档
PUT goods_index/_doc/1
{
"name":"张三1111",
"age":34,
"desc":"北京人深圳来"
}
# 删除文档
DELETE goods_index/_doc/1
# 查询文档 指定id
GET goods_index/_doc/56YoJ5IBpXr2Jk-kqx2q
# 查询所有文档
GET goods_index/_search
# 删除索引
DELETE goods_index
五、向量数据测试，kibana脚本，操作es
1. 创建索引
PUT /vector_index
{
"mappings": {
"properties": {
"my_vector_field": {
"type": "dense_vector",
"dims": 128
},
"my_text": {
"type": "text"
}
}
}
}
2. 写入数据
PUT /vector_index/_doc/1
{
"my_text": "示例文本",
"my_vector_field": [0.1, 0.2, ..., 0.128]
}
3. 搜索查询
POST /vector_index/_search
{
"query": {
"script_score": { // 使用脚本评分来计算相似度得分
"query": {
"match_all": {}
},
"script": {
"source": "cosineSimilarity(params.queryVector, 'my_vector_field') + 1.0", // 计算查询向量与存储向量的余弦相似度，并加 1.0 以确保得分非负
"params": {
"queryVector": [0.1, 0.2, ..., 0.128]
}
}
}
}
}
参考资料
-
https://blog.csdn.net/cmh1008611/article/details/141562198
-
https://blog.csdn.net/qq_50854662/article/details/135967448
开源项目地址：https://gitee.com/s11e-DAO/bsin-paas-os
原文地址：https://blog.csdn.net/leijiwen/article/details/142522475
免责声明：本站文章内容转载自网络资源，如本站内容侵犯了原著者的合法权益，可联系本站删除。更多内容请关注自学内容网（zxcms.com）！