diff --git a/.github/ISSUE_TEMPLATE/workflows/book.yml b/.github/ISSUE_TEMPLATE/workflows/book.yml
new file mode 100644
index 0000000..55e548d
--- /dev/null
+++ b/.github/ISSUE_TEMPLATE/workflows/book.yml
@@ -0,0 +1,32 @@
+name: Book
+
+on:
+  push:
+    branches:
+      - release
+    paths:
+      - 'docs/**'
+
+jobs:
+  deploy_en:
+    name: Deploy book on gh-pages
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2
+      - name: Install mdBook
+        uses: peaceiris/actions-mdbook@v1
+      - name: Render book
+        run: |
+          mdbook build -d gh-pages docs/zh-CN
+          mkdir gh-pages
+          mv docs/zh-CN/gh-pages gh-pages/zh-CN
+      - name: Deploy
+        uses: peaceiris/actions-gh-pages@v2.5.1
+        with:
+          emptyCommits: true
+          keepFiles: false
+        env:
+          ACTIONS_DEPLOY_KEY: ${{ secrets.ACTIONS_DEPLOY_KEY }}
+          PUBLISH_BRANCH: gh-pages
+          PUBLISH_DIR: gh-pages
\ No newline at end of file
diff --git a/.github/workflows/book.yml b/.github/workflows/book.yml
new file mode 100644
index 0000000..ff3f0c9
--- /dev/null
+++ b/.github/workflows/book.yml
@@ -0,0 +1,34 @@
+name: Book
+
+on:
+  push:
+    branches:
+      - release
+    paths:
+      - 'docs/**'
+
+jobs:
+  deploy_en:
+    name: Deploy book on gh-pages
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2
+      - name: Install mdBook
+        uses: peaceiris/actions-mdbook@v1
+      - name: Render book
+        run: |
+          mdbook build -d gh-pages docs/zh-CN
+          mkdir gh-pages
+          mv docs/zh-CN/gh-pages gh-pages/zh-CN
+          mv docs/index.html gh-pages
+
+      - name: Deploy
+        uses: peaceiris/actions-gh-pages@v2.5.1
+        with:
+          emptyCommits: true
+          keepFiles: false
+        env:
+          ACTIONS_DEPLOY_KEY: ${{ secrets.ACTIONS_DEPLOY_KEY }}
+          PUBLISH_BRANCH: gh-pages
+          PUBLISH_DIR: gh-pages
diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
new file mode 100644
index 0000000..88bbf62
--- /dev/null
+++ b/.github/workflows/release.yml
@@ -0,0 +1,39 @@
+
+
+on:
+  push:
+    tags:
+    - 'v*' 
+
+name: release
+
+jobs:
+  build:
+    name: Upload Release Asset
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v2
+      - name: Build project # This would actually build your project, using zip for an example artifact
+        run: |
+          cargo build --release
+      - name: Create Release
+        id: create_release
+        uses: actions/create-release@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.ACTIONS_DEPLOY_KEY }}
+        with:
+          tag_name: ${{ github.ref }}
+          release_name: Release ${{ github.ref }}
+          draft: false
+          prerelease: false
+      - name: Upload Release Asset
+        id: upload-release-asset 
+        uses: actions/upload-release-asset@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.ACTIONS_DEPLOY_KEY }}
+        with:
+          upload_url: ${{ steps.create_release.outputs.upload_url }} # This pulls from the CREATE RELEASE step above, referencing it's ID to get its outputs object, which include a `upload_url`. See this blog post for more info: https://jasonet.co/posts/new-features-of-github-actions/#passing-data-to-future-steps 
+          asset_path: ./target/release/chubaodb
+          asset_name: chubaodb-linux
+          asset_content_type: application/zip
diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml
new file mode 100644
index 0000000..3c13d1b
--- /dev/null
+++ b/.github/workflows/rust.yml
@@ -0,0 +1,22 @@
+name: Rust
+
+on:
+  push:
+    branches: [ master ]
+  pull_request:
+    branches: [ master ]
+
+env:
+  CARGO_TERM_COLOR: always
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Build
+      run: cargo build --verbose
+    - name: Run tests
+      run: cargo test --verbose
diff --git a/README.md b/README.md
index 7261a86..772d538 100644
--- a/README.md
+++ b/README.md
@@ -11,6 +11,7 @@ As a scalable non-relational structured data infrastructure, ChubaoDB has severa
 * rich indexes for efficient search
 
 
+
 ## External Interface
 
 collection, document, field
diff --git a/docs/index.html b/docs/index.html
new file mode 100644
index 0000000..ddf77d6
--- /dev/null
+++ b/docs/index.html
@@ -0,0 +1 @@
+<a href="zh-CN/"> 中文文档 </a>
\ No newline at end of file
diff --git a/docs/zh-CN/.gitignore b/docs/zh-CN/.gitignore
new file mode 100644
index 0000000..7585238
--- /dev/null
+++ b/docs/zh-CN/.gitignore
@@ -0,0 +1 @@
+book
diff --git a/docs/zh-CN/book.toml b/docs/zh-CN/book.toml
new file mode 100644
index 0000000..db0559d
--- /dev/null
+++ b/docs/zh-CN/book.toml
@@ -0,0 +1,6 @@
+[book]
+authors = ["AnsjSun"]
+language = "cn"
+multilingual = false
+src = "src"
+title = "chubaodb 白皮书"
diff --git a/docs/zh-CN/src/SUMMARY.md b/docs/zh-CN/src/SUMMARY.md
new file mode 100644
index 0000000..3757e45
--- /dev/null
+++ b/docs/zh-CN/src/SUMMARY.md
@@ -0,0 +1,16 @@
+# Summary
+
+- [介绍](./introduction.md)
+- [编译与安装](./install.md)
+    - [配置文件](./config.md)
+    - [集群模式](./cluster.md)
+- [元数据管理](./master.md)
+    - [库表管理](./collection.md)
+- [数据操作](./document.md)
+    - [CRUD](./crud.md)
+    - [搜索](./search.md)
+    - [聚合](./aggregation.md)
+    - [向量](./vector.md)
+- [设计](./design.md)
+    - [架构](./architecture.md)
+    - [计划](./plan.md)
diff --git a/docs/zh-CN/src/aggregation.md b/docs/zh-CN/src/aggregation.md
new file mode 100644
index 0000000..a2c89c1
--- /dev/null
+++ b/docs/zh-CN/src/aggregation.md
@@ -0,0 +1,127 @@
+# 聚合那些事儿
+
+下面我们来介绍一下chubaodb的聚合功能，在这之前，你确信已经通过[库表管理](*./collection.md*)创建了表。
+
+我们先插入一些测试数据吧，先创建5个人
+
+````
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "张三",
+	"age": 20,
+	"birthday": "2000-02-02",
+	"description": "zhangsan can use java good at php, but python not good at",
+	"skills": ["java", "php", "python"]
+}
+' "http://127.0.0.1:8080/put/person/1"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "李四",
+	"age": 30,
+	"birthday": "1990-02-02",
+	"description": "lisi can use java ,only java",
+	"skills": ["java"]
+}
+' "http://127.0.0.1:8080/put/person/2"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "王五",
+	"age": 20,
+	"birthday": "2000-03-20",
+	"description": "wangwu is c++ rust good at!",
+	"skills": ["c++", "rust"]
+}
+' "http://127.0.0.1:8080/put/person/3"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "牛六",
+	"age": 35,
+	"birthday": "1985-12-02",
+	"description": "niuliu age too old",
+	"skills": ["java", "golang", "python", "rust"]
+}
+' "http://127.0.0.1:8080/put/person/4"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "赵七",
+	"age": 18,
+	"birthday": "2002-03-12",
+	"description": "is a web user, he can use ruby and php",
+	"skills": ["php", "ruby"]
+}
+' "http://127.0.0.1:8080/put/person/5"
+````
+
+
+
+插入完成后，我们通过search接口可以查看到这五个人 `http://127.0.0.1:8080/search/person`
+
+
+
+我们先了解下聚合的真谛。 我们从一句大家都熟悉的sql 入手吧
+
+`select age,count(age) from person group by age`
+
+这个聚合函数可以堪称两个部分。分别是 `group by ` 分组，和 `count(*)` 指标连个部分。所以聚合份两种，一个是分组，一个和统计。
+
+ok，现在让我们看看在`chubaodb`中这两部分是如何完成的。我们通过一个例子入手
+
+`http://127.0.0.1:8080/agg/person?group=term(age)&fun=stats(age)`
+
+* agg 是方法的操作
+* group= 是聚合方法。例子中我们用的term 方式进行分组。
+* fun = 是指标方法。例子中我们统计 max ，min 等
+
+![image-20200715135803824](image/image-20200715135803824.png)
+
+
+
+好的恭喜你已经看到结果了。默认是按照value中的count 进行排序
+
+
+
+### 目前group 支持的方法有
+
+`term(name)` 按照字段进行聚合
+
+`date(name,format)` example: date(birthday, yyyy-MM) 按照时间格式进行聚合
+
+`range(name, range_str)` example:*range(age,-0,0-20,20-30,30-40,40-)*
+
+
+
+### Fun 支持的方法有
+
+`hits(size)`:每个分组中返回size个文档，无排序功能
+
+`stats(name)`: 每个分组中，count max, min , sum missing 的个数
+
+* fun 可以为空，如果为空则默认为count方式
+
+
+
+现在让我们自举一些例子：
+
+### 按照年龄分组，分别显示每个年龄中的人
+
+  ![image-20200715140655966](image/image-20200715140655966.png)
+
+### 照技能分组，显示每个技能的人数
+
+`http://127.0.0.1:8080/agg/person?group=term(skills)`
+
+![image-20200715140812082](image/image-20200715140812082.png)
+
+### 按照技能分组，显示每个技能的人数,按照技能名称排序
+
+`http://127.0.0.1:8080/agg/person?group=term(skills)&sort=key:asc`
+
+![image-20200715140902639](image/image-20200715140902639.png)
+
+### 按照年龄和技能分组查看人数
+
+![image-20200715141018443](image/image-20200715141018443.png)
\ No newline at end of file
diff --git a/docs/zh-CN/src/architecture.md b/docs/zh-CN/src/architecture.md
new file mode 100644
index 0000000..c503cff
--- /dev/null
+++ b/docs/zh-CN/src/architecture.md
@@ -0,0 +1 @@
+# 关于架构
diff --git a/docs/zh-CN/src/cluster.md b/docs/zh-CN/src/cluster.md
new file mode 100644
index 0000000..373f8f6
--- /dev/null
+++ b/docs/zh-CN/src/cluster.md
@@ -0,0 +1,3 @@
+# 集群模式
+
+编造ing......
diff --git a/docs/zh-CN/src/collection.md b/docs/zh-CN/src/collection.md
new file mode 100644
index 0000000..30ccbe7
--- /dev/null
+++ b/docs/zh-CN/src/collection.md
@@ -0,0 +1,60 @@
+# 库表管理
+
+下面我们会对`创建表`,`查询表`，`删除表` 进行演示。目前表结构一旦创建不支持修改。
+打开master的管理地址 http://127.0.0.1:7070
+
+### 创建表
+
+通过 `collectionCreate`接口进行创建
+
+![image-20200715112907536](image/image-20200715112907536.png)
+
+* Name 是表名称， 
+* partitionNum 是这个表分多少个分片。分片多会提高插入的并发能力，但是会降低搜索效率，并非越多或者越少越好
+* partitionReplicaNum 是每个分片多少个副本。建议要么1，要么3+ 。在传统分布式系统环境，可以设置为3，单机版智能设置1.partitionReplicaNum 必须小于等于你的机器个数
+* Fields 是这个表里面的字段。我们提供了 `int`, `float`, `string`, `text`, `vector`, `date` 几种字段格式，注意 text 和string的区别是。text是全文检索，比如 `中国银行` 搜索`中国`是会被召回的， `string`的话必须输入完整的 匹配。
+
+
+
+下面我们创建一个人物表，包含 `ID`， `姓名`， `年龄` ，`生日` ， `简介` ，`技能` 几个字段，
+
+````
+mutation{
+  collectionCreate(
+    name:"person", 
+    partitionNum:1, 
+    partitionReplicaNum:1
+  	fields:{
+      int:[{name:"age", value:true }]
+      string:[{name:"name" }, {name:"skills", value:true, array:true}]
+      date:[{name:"birthday", value:true }]
+      text:[{name:"description"}]
+    }
+  
+  )
+}
+````
+
+![image-20200715115437856](image/image-20200715115437856.png)
+
+
+
+出现如下结构意味着创建表成功了。每种类型有自己的参数大家可以参阅iql的文档。
+
+
+
+### 查询表
+
+![image-20200715115617185](image/image-20200715115617185.png)
+
+
+
+
+
+删除表
+
+![image-20200715115703796](image/image-20200715115703796.png)
+
+
+
+ok 你已经具备了元数据管理的基本技能。
\ No newline at end of file
diff --git a/docs/zh-CN/src/config.md b/docs/zh-CN/src/config.md
new file mode 100644
index 0000000..f650080
--- /dev/null
+++ b/docs/zh-CN/src/config.md
@@ -0,0 +1,156 @@
+# 配置文件
+
+前面我们已经学会了，如果以最基本的方式启动chubaodb。下面我们来学一种非常高级的用法，通过配置文件和启动参数来启动chubaodb。在学这个之前我们可能需要了简单解下chubaodb的架构，chubaodb份了三个角色`master`,`pserver`,`router`.
+
+* master 是集群管理，元数据管理的模块。表结构，数据均衡，failover 等
+* pserver 是数据存储及计算节点。
+* router 是提供了restful的数据查询组件。router 是无状态的。
+
+下面是一个简单的config例子。也是默认config的参数
+
+````
+[global]
+# the name will validate join cluster by same name
+name = "chubaodb"
+# your server ip for connect, you can use -i in rags to set it
+# ip = "127.0.0.1"
+# log path , If you are in a production environment, You'd better set absolute paths
+log = "logs/"
+# default log type for any model
+log_level = "info"
+# log file size for rolling
+log_limit_bytes = 128000000
+# number of reserved log files
+log_file_count = 10
+# Whether to use distributed storage. If it's true, there's only one
+shared_disk = false
+
+[router]
+# port for server
+http_port = 8080
+
+[ps]
+#set zone num, default is 0
+zone = "default"
+# you data save to disk path ,If you are in a production environment, You'd better set absolute paths
+data = "data/"
+# port for server
+rpc_port = 9090
+# how often to refresh the index
+flush_sleep_sec = 3
+    [ps.raft]
+        heartbeat_port = 10030
+        replicate_port = 10031
+        # how size of num for memory
+        log_max_num = 200000
+        # how size of num for memory
+        log_min_num = 100000
+        # how size of num for memory
+        log_file_size_mb = 128
+        # Three  without a heartbeat , follower to begin consecutive elections
+        heartbeate_ms = 500
+    
+
+[[masters]]
+# master ip for service
+ip = "127.0.0.1"
+# port for server
+http_port = 7070
+# master data path for meta
+data = "data/meta/"
+````
+
+可以发现我们把 三个模块写到了同一个配置文件。同时各个节点通过参数来选择启动的模块。实现了一个配置文件走天下的易用功能。
+
+可以通过 `./chubaodb --help` 获得参数说明
+
+````
+(base) ➜  release git:(async-std) ✗ ./chubaodb --help
+chubaodb 0.1.0
+hello index world
+
+USAGE:
+    chubaodb [SUBCOMMAND]
+
+FLAGS:
+    -h, --help       Prints help information
+    -V, --version    Prints version information
+
+SUBCOMMANDS:
+    all       
+    help      Prints this message or the help of the given subcommand(s)
+    master    
+    ps        
+    router   
+````
+
+启动可以通过 `./chubaodb all --help ` , `./chubaodb ps --help ` .....来获取更多。参数说明
+
+* 比如我们想把三个模块在一个进程中启动 那么就是. `./chubaodb all -c ../../config/config.toml` .
+
+````
+./chubaodb all -c ../../config/config.toml
+load config by path: ../../config/config.toml
+2020-07-15 11:05:39 - INFO - chubaodb::util::config(189) - log init ok 
+2020-07-15 11:05:39 - INFO - chubaodb(149) - All ChubaoDB servers were started successfully!
+2020-07-15 11:05:39 - INFO - chubaodb::router::server(29) - router is listening on http://0.0.0.0:8080
+2020-07-15 11:05:39 - INFO - actix_server::builder(262) - Starting 8 workers
+2020-07-15 11:05:39 - INFO - chubaodb::util::http_client(21) - send get for url:http://127.0.0.1:7070/my_ip
+2020-07-15 11:05:39 - INFO - surf::middleware::logger::native(119) - sending request
+2020-07-15 11:05:39 - INFO - actix_server::builder(276) - Starting "actix-web-service-0.0.0.0:8080" service on 0.0.0.0:8080
+2020-07-15 11:05:39 - WARN - isahc::handler(209) - request completed with error [id=AtomicCell { value: 0 }]: ConnectFailed: failed to connect to the server
+2020-07-15 11:05:39 - ERROR - chubaodb::pserver::server(41) - got ip from master has err:Error(InternalErr, "ConnectFailed: failed to connect to the server")
+2020-07-15 11:05:39 - INFO - chubaodb::master::server(43) - master listening on http://0.0.0.0:7070
+2020-07-15 11:05:39 - INFO - actix_server::builder(262) - Starting 8 workers
+2020-07-15 11:05:39 - INFO - actix_server::builder(276) - Starting "actix-web-service-0.0.0.0:7070" service on 0.0.0.0:7070
+2020-07-15 11:05:40 - INFO - chubaodb::util::http_client(21) - send get for url:http://127.0.0.1:7070/my_ip
+2020-07-15 11:05:40 - INFO - surf::middleware::logger::native(119) - sending request
+2020-07-15 11:05:40 - INFO - chubaodb::master::server(440) - success_response [Object({"ip": String("127.0.0.1")})]
+2020-07-15 11:05:40 - INFO - surf::middleware::logger::native(119) - request completed
+2020-07-15 11:05:40 - INFO - chubaodb::pserver::server(36) - got my ip:127.0.0.1 from master
+2020-07-15 11:05:40 - INFO - chubaodb::util::http_client(51) - send post for url:http://127.0.0.1:7070/pserver/register
+2020-07-15 11:05:40 - INFO - surf::middleware::logger::native(119) - sending request
+2020-07-15 11:05:40 - INFO - chubaodb::master::server(301) - prepare to heartbeat with address 127.0.0.1:9090, zone default
+2020-07-15 11:05:40 - INFO - chubaodb::master::server(440) - success_response [PServer { id: Some(1), addr: "127.0.0.1:9090", write_partitions: [], zone: "default", modify_time: 0 }]
+2020-07-15 11:05:40 - INFO - surf::middleware::logger::native(119) - request completed
+2020-07-15 11:05:40 - INFO - chubaodb::pserver::service(111) - register to master ok: node_id:Some(1) 
+2020-07-15 11:05:40 - INFO - chubaodb::pserver::service(126) - register server line:PServer { id: Some(1), addr: "127.0.0.1:9090", write_partitions: [], zone: "default", modify_time: 0 }
+2020-07-15 11:05:40 - INFO - chubaodb::pserver::server(59) - init pserver OK use time:Ok(5.333ms)
+````
+
+* 比如我们只启动 router 那么就是. `./chubaodb router -c ../../config/config.toml` .
+
+````
+(base) ➜  release git:(async-std) ✗ ./chubaodb router -c ../../config/config.toml
+load config by path: ../../config/config.toml
+2020-07-15 11:01:59 - INFO - chubaodb::util::config(189) - log init ok 
+2020-07-15 11:01:59 - INFO - chubaodb(149) - All ChubaoDB servers were started successfully!
+2020-07-15 11:01:59 - INFO - chubaodb::router::server(29) - router is listening on http://0.0.0.0:8080
+2020-07-15 11:01:59 - INFO - actix_server::builder(262) - Starting 8 workers
+2020-07-15 11:01:59 - INFO - actix_server::builder(276) - Starting "actix-web-service-0.0.0.0:8080" service on 0.0.0.0:8080
+````
+
+* 比如我们只启动 master 那么就是. `./chubaodb master -c ../../config/config.toml` .
+
+````
+load config by path: ../../config/config.toml
+2020-07-15 11:03:11 - INFO - chubaodb::util::config(189) - log init ok 
+2020-07-15 11:03:11 - INFO - chubaodb(149) - All ChubaoDB servers were started successfully!
+2020-07-15 11:03:11 - INFO - chubaodb::master::server(43) - master listening on http://0.0.0.0:7070
+2020-07-15 11:03:11 - INFO - actix_server::builder(262) - Starting 8 workers
+2020-07-15 11:03:11 - INFO - actix_server::builder(276) - Starting "actix-web-service-0.0.0.0:7070" service on 0.0.0.0:7070
+^C2020-07-15 11:03:12 - INFO - actix_server::builder(321) - SIGINT received, exiting
+````
+
+
+* 比如我们只启动 pserver 那么就是. `./chubaodb ps -c ../../config/config.toml` . ps会通过你配置文件中master的配置自动去master 注册为数据节点
+
+````
+load config by path: ../../config/config.toml
+2020-07-15 11:03:45 - INFO - chubaodb::util::config(189) - log init ok 
+2020-07-15 11:03:45 - INFO - chubaodb(149) - All ChubaoDB servers were started successfully!
+2020-07-15 11:03:45 - INFO - chubaodb::util::http_client(21) - send get for url:http://127.0.0.1:7070/my_ip
+2020-07-15 11:03:45 - INFO - surf::middleware::logger::native(119) - sending request
+2020-07-15 11:03:45 - WARN - isahc::handler(209) - request completed with error [id=AtomicCell { value: 0 }]: ConnectFailed: failed to connect to the server
+````
+
diff --git a/docs/zh-CN/src/crud.md b/docs/zh-CN/src/crud.md
new file mode 100644
index 0000000..88451de
--- /dev/null
+++ b/docs/zh-CN/src/crud.md
@@ -0,0 +1,74 @@
+# crud那些事儿
+
+在[库表管理](./collection.md)里我们学习了如何创建一个表。下面让我们对这个表进行数据的增删改查。
+
+在这之前我们了解下插入数据的姿势。
+
+* `put` 代表不管有没有都插进去。document的version归为1
+* `create` 必须没有，如果有报错，document的version为1
+* `update` 必须存在，如果不存在就报错， document的version 递增+1
+* `upsert` 有则走`update`逻辑，没有则走`create`逻辑
+
+好了你已经学会了存储的真谛，让我们来试着插入一条数据吧！注意数据操作是在 router 上进行，也就是默认的`8080`端口上。
+
+## put
+
+````
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "张三",
+	"age": 20,
+	"birthday": "2000-02-02",
+	"description": "一个用来测试的倒霉蛋",
+	"skills": ["java", "php", "python"]
+}
+' "http://127.0.0.1:8080/put/person/1"
+````
+看到如下
+
+![image-20200715124902323](image/image-20200715124902323.png)
+
+代表插入成功！
+
+* `http://127.0.0.1:8080/put/person/1` 地址中`person` 是我们创建的表名称， 1 位当前用户的唯一id。字符串格式。（ps:id 还有一种方式，双keyid。这是一种很高级的做法，后门我们会对次情况单门一章来说明）
+
+我们可以通过get 接口来获取这条数据 http://127.0.0.1:8080/get/person/1
+
+
+
+![image-20200715125145828](image/image-20200715125145828.png)
+
+## update
+
+比如我们尝试更新张三的技能增加rust 通过如下方式
+
+````
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"skills": ["java", "php", "python","rust"]
+}
+' "http://127.0.0.1:8080/update/person/1"
+````
+
+
+
+![image-20200715125335276](image/image-20200715125335276.png)
+
+我们可以开心的看到张三学会了`rust` 并且 version 改为了2.
+
+
+
+## delete
+
+通过 
+
+```
+curl -XDELETE http://127.0.0.1:8080/delete/person/1
+{"code":200,"message":"success"}
+```
+
+可以删除张三这条记录。我们看看删除后再做get会得到什么
+
+![image-20200715125658865](image/image-20200715125658865.png)
+
+嗯。很好本章结束了！
\ No newline at end of file
diff --git a/docs/zh-CN/src/design.md b/docs/zh-CN/src/design.md
new file mode 100644
index 0000000..925a8be
--- /dev/null
+++ b/docs/zh-CN/src/design.md
@@ -0,0 +1 @@
+# 设计
diff --git a/docs/zh-CN/src/document.md b/docs/zh-CN/src/document.md
new file mode 100644
index 0000000..b7e43d8
--- /dev/null
+++ b/docs/zh-CN/src/document.md
@@ -0,0 +1 @@
+# 数据操作
diff --git a/docs/zh-CN/src/image/image-20200715111140350.png b/docs/zh-CN/src/image/image-20200715111140350.png
new file mode 100644
index 0000000..f1717ed
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715111140350.png differ
diff --git a/docs/zh-CN/src/image/image-20200715111454748.png b/docs/zh-CN/src/image/image-20200715111454748.png
new file mode 100644
index 0000000..570ff47
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715111454748.png differ
diff --git a/docs/zh-CN/src/image/image-20200715111749876.png b/docs/zh-CN/src/image/image-20200715111749876.png
new file mode 100644
index 0000000..58fc9fd
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715111749876.png differ
diff --git a/docs/zh-CN/src/image/image-20200715111932528.png b/docs/zh-CN/src/image/image-20200715111932528.png
new file mode 100644
index 0000000..2c122b1
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715111932528.png differ
diff --git a/docs/zh-CN/src/image/image-20200715112907536.png b/docs/zh-CN/src/image/image-20200715112907536.png
new file mode 100644
index 0000000..aaf99de
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715112907536.png differ
diff --git a/docs/zh-CN/src/image/image-20200715115437856.png b/docs/zh-CN/src/image/image-20200715115437856.png
new file mode 100644
index 0000000..d98bfb2
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715115437856.png differ
diff --git a/docs/zh-CN/src/image/image-20200715115617185.png b/docs/zh-CN/src/image/image-20200715115617185.png
new file mode 100644
index 0000000..e72b520
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715115617185.png differ
diff --git a/docs/zh-CN/src/image/image-20200715115703796.png b/docs/zh-CN/src/image/image-20200715115703796.png
new file mode 100644
index 0000000..9f5c3f8
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715115703796.png differ
diff --git a/docs/zh-CN/src/image/image-20200715124902323.png b/docs/zh-CN/src/image/image-20200715124902323.png
new file mode 100644
index 0000000..7c440ce
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715124902323.png differ
diff --git a/docs/zh-CN/src/image/image-20200715125145828.png b/docs/zh-CN/src/image/image-20200715125145828.png
new file mode 100644
index 0000000..5aee525
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715125145828.png differ
diff --git a/docs/zh-CN/src/image/image-20200715125335276.png b/docs/zh-CN/src/image/image-20200715125335276.png
new file mode 100644
index 0000000..195c9cd
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715125335276.png differ
diff --git a/docs/zh-CN/src/image/image-20200715125349302.png b/docs/zh-CN/src/image/image-20200715125349302.png
new file mode 100644
index 0000000..195c9cd
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715125349302.png differ
diff --git a/docs/zh-CN/src/image/image-20200715125658865.png b/docs/zh-CN/src/image/image-20200715125658865.png
new file mode 100644
index 0000000..e2dd0f7
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715125658865.png differ
diff --git a/docs/zh-CN/src/image/image-20200715130811821.png b/docs/zh-CN/src/image/image-20200715130811821.png
new file mode 100644
index 0000000..057980d
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715130811821.png differ
diff --git a/docs/zh-CN/src/image/image-20200715132053132.png b/docs/zh-CN/src/image/image-20200715132053132.png
new file mode 100644
index 0000000..c10cdcf
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715132053132.png differ
diff --git a/docs/zh-CN/src/image/image-20200715134112636.png b/docs/zh-CN/src/image/image-20200715134112636.png
new file mode 100644
index 0000000..741451c
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715134112636.png differ
diff --git a/docs/zh-CN/src/image/image-20200715134143880.png b/docs/zh-CN/src/image/image-20200715134143880.png
new file mode 100644
index 0000000..3df6e66
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715134143880.png differ
diff --git a/docs/zh-CN/src/image/image-20200715134337378.png b/docs/zh-CN/src/image/image-20200715134337378.png
new file mode 100644
index 0000000..ff90e2d
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715134337378.png differ
diff --git a/docs/zh-CN/src/image/image-20200715134556712.png b/docs/zh-CN/src/image/image-20200715134556712.png
new file mode 100644
index 0000000..de8c01d
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715134556712.png differ
diff --git a/docs/zh-CN/src/image/image-20200715134811989.png b/docs/zh-CN/src/image/image-20200715134811989.png
new file mode 100644
index 0000000..6168f27
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715134811989.png differ
diff --git a/docs/zh-CN/src/image/image-20200715135803824.png b/docs/zh-CN/src/image/image-20200715135803824.png
new file mode 100644
index 0000000..ef66529
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715135803824.png differ
diff --git a/docs/zh-CN/src/image/image-20200715140655966.png b/docs/zh-CN/src/image/image-20200715140655966.png
new file mode 100644
index 0000000..4b388bb
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715140655966.png differ
diff --git a/docs/zh-CN/src/image/image-20200715140812082.png b/docs/zh-CN/src/image/image-20200715140812082.png
new file mode 100644
index 0000000..48b414c
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715140812082.png differ
diff --git a/docs/zh-CN/src/image/image-20200715140902639.png b/docs/zh-CN/src/image/image-20200715140902639.png
new file mode 100644
index 0000000..97a6c67
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715140902639.png differ
diff --git a/docs/zh-CN/src/image/image-20200715141018443.png b/docs/zh-CN/src/image/image-20200715141018443.png
new file mode 100644
index 0000000..10a6194
Binary files /dev/null and b/docs/zh-CN/src/image/image-20200715141018443.png differ
diff --git a/docs/zh-CN/src/install.md b/docs/zh-CN/src/install.md
new file mode 100644
index 0000000..80d449e
--- /dev/null
+++ b/docs/zh-CN/src/install.md
@@ -0,0 +1,32 @@
+# 编译与安装
+
+`chubaodb`采用rust编写同时可能会依赖一些c库，支持跨平台编译，目前支持操作系统有 windows , linux , macos.可能还有其他不知道的，碰到再说。
+
+
+
+|  共功能\系统   | windows  | linux  | macos  |
+|  ----  | ----  | ----  | ----  |
+| 存储  | 支持 | 支持 | 支持 |
+| 全文检索  | 支持 | 支持 | 支持 |
+| 向量检索  | `不支持` | 支持 | 支持 |
+
+
+-----------------
+
+**重点**: 如果你之前没有rust环境，且对rust 没什么兴趣直接下载我们的release 版本。 (TODO: release连接)，珍爱生命远离rust编译
+
+
+**开始编译以在macos 下编译为例**
+* 在项目根目录执行 `cargo build --release` ，
+* 如果运气好没有报错，那么进入 `target/release` 目录获得编译好的二进制文件 `chubaodb`
+* 执行 ./chubaodb 
+* 打开浏览器访问 http://127.0.0.1:8080 为router地址， 这个地址用来对数据进行CRUD
+* 打开浏览器访问 http://127.0.0.1:7070 为master地址， 这个地址用来对库表及数据结构进行管理。
+* 之后的章节会详细介绍这些地址的用法，如果一切顺利恭喜你，chubaodb已经在你的电脑中完美运行了。
+
+
+
+
+
+
+
diff --git a/docs/zh-CN/src/introduction.md b/docs/zh-CN/src/introduction.md
new file mode 100644
index 0000000..053c2c0
--- /dev/null
+++ b/docs/zh-CN/src/introduction.md
@@ -0,0 +1,16 @@
+
+### 介绍
+`chubaodb` 是一个分布式高可用的云原生，同时支持传统的分布式文档搜索及存储系统，支持`全文检索`,`聚合查询`,`向量搜索`,`标量搜索`的功能，采用轻schema策略，尽可能提高了存储文档的灵活度。同时吸取其他类似软件的经验，初心于在有限的计算节点情况下，支持不限容量的存储及计算，同时尽可能低的学习成本，完成尽可能多的需求。
+
+### 初心
+我们期望于`chubaodb` 是一个一看就会，开箱即用的软件，让使用者不在对被使用的软件所使用。严格准守，山寨三体的软件设计原则， 
+* 1.程序必须正确 
+* 2.程序必须可维护，如果和第1条冲突，以第1条为准 
+* 3.程序必须高效，如和第1条冲突，则以第1条为准，如和第2条冲突，则以第2条为准。
+
+### 特点
+* `graphql` 进行schema管理， 
+* `restful api`进行数据操作， 
+* `raft`保证数据的最终一致性， 
+* `rocksdb`, `tantivy`， `faiss` 提供了底层存储及计算的能力。
+* 通过`chubaofs` 提供了无限存储的可能，通过架构可以释放或加载宝贵的cpu和内存资源。
\ No newline at end of file
diff --git a/docs/zh-CN/src/master.md b/docs/zh-CN/src/master.md
new file mode 100644
index 0000000..683b413
--- /dev/null
+++ b/docs/zh-CN/src/master.md
@@ -0,0 +1,45 @@
+# 元数据管理
+
+元数据管理是在master api上，提供了graphql的方式。同时内置了一个iql。如果你没有对配置做过更改，master 的地址应该为 http://127.0.0.1:7070
+
+打开地址你就看到如下
+
+![image-20200715111140350](image/image-20200715111140350.png)
+
+
+
+熟悉iql的就不解释了，不熟悉的用用就差不多。
+
+
+
+我们简单说一下各个接口的功能，以collectionGet 为例，这个接口是获取一个表的结构， 点击右侧第二个方法。可以看到需要的参数
+
+![image-20200715111454748](image/image-20200715111454748.png)
+
+
+
+参数为，id ，name 。类型后门没有跟`!`意思就非必须字段。但是必须二选一。意味着你可以通过一个name 或者id 去查询这个collection.然后我们输入query
+
+````
+{
+  collectionGet(name:"t1")
+}
+````
+
+
+
+
+
+![image-20200715111749876](image/image-20200715111749876.png)
+
+
+
+
+
+可以看到报错了。没错。就是错了。因为我没还没有创建名字为`t1`的collection。后门会有创建的方式。那我们换一个简单的的吧。我们通过query来查询当前系统的版本。
+
+![image-20200715111932528](image/image-20200715111932528.png)
+
+
+
+红色部分为返回结果。太过简单就不解释了。
\ No newline at end of file
diff --git a/docs/zh-CN/src/plan.md b/docs/zh-CN/src/plan.md
new file mode 100644
index 0000000..94322c0
--- /dev/null
+++ b/docs/zh-CN/src/plan.md
@@ -0,0 +1 @@
+# 计划
diff --git a/docs/zh-CN/src/search.md b/docs/zh-CN/src/search.md
new file mode 100644
index 0000000..306265b
--- /dev/null
+++ b/docs/zh-CN/src/search.md
@@ -0,0 +1,109 @@
+# 搜索那些事儿
+
+下面我们来介绍一下chubaodb的搜索功能，在这之前，你确信已经通过[库表管理](*./collection.md*)创建了表。
+
+我们先插入一些测试数据吧，先创建5个人
+
+````
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "张三",
+	"age": 20,
+	"birthday": "2000-02-02",
+	"description": "zhangsan can use java good at php, but python not good at",
+	"skills": ["java", "php", "python"]
+}
+' "http://127.0.0.1:8080/put/person/1"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "李四",
+	"age": 30,
+	"birthday": "1990-02-02",
+	"description": "lisi can use java ,only java",
+	"skills": ["java"]
+}
+' "http://127.0.0.1:8080/put/person/2"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "王五",
+	"age": 20,
+	"birthday": "2000-03-20",
+	"description": "wangwu is c++ rust good at!",
+	"skills": ["c++", "rust"]
+}
+' "http://127.0.0.1:8080/put/person/3"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "牛六",
+	"age": 35,
+	"birthday": "1985-12-02",
+	"description": "niuliu age too old",
+	"skills": ["java", "golang", "python", "rust"]
+}
+' "http://127.0.0.1:8080/put/person/4"
+
+curl -H "Content-Type: application/json" -XPOST -d'
+{
+	"name": "赵七",
+	"age": 18,
+	"birthday": "2002-03-12",
+	"description": "is a web user, he can use ruby and php",
+	"skills": ["php", "ruby"]
+}
+' "http://127.0.0.1:8080/put/person/5"
+````
+
+
+
+插入完成后，我们通过search接口可以查看到这五个人 `http://127.0.0.1:8080/search/person`
+
+
+
+![image-20200715134112636](image/image-20200715134112636.png)
+
+
+
+http://127.0.0.1:8080/search/person` 参数为空时语义为query:* search 又如下参数
+
+*  query 查询语句，也就是类似lucene 的dsl 。（TODO： 专门一章介绍dsl）
+*  def_fields 默认查询字段。当query不指定字段时候以此字段为查询，为or的关系，可以多个字段用逗号`,`隔开
+*  size: 返回数据条数，默认为20
+*  sort: 排序规则 example：*name:asc|age:desc* , 默认为score排序也就是相关度
+
+下面我们把这些query 都用上做一个查询吧！
+
+`http://127.0.0.1:8080/search/person?query=name:%E5%BC%A0%E4%B8%89&size=3&sort=age:asc`
+
+![image-20200715134143880](image/image-20200715134143880.png)
+
+下面让我们自举一些需求。
+
+查找摘要中 包含 rust 或者 java 的人 
+
+* `http://127.0.0.1:8080/search/person?query=description:java%20OR%20description:rust&size=3&sort=age:asc`
+* 上述语句等同于 `http://127.0.0.1:8080/search/person?query=java%20OR%20rust&def_fields=description&size=3&sort=age:asc`
+
+
+
+![image-20200715134337378](image/image-20200715134337378.png)
+
+
+
+查找摘要中 包含  java 的人  按照年龄倒序
+
+````
+http://127.0.0.1:8080/search/person?query=java&def_fields=description&size=3&sort=age:desc
+````
+
+![image-20200715134556712](image/image-20200715134556712.png)
+
+
+
+### 精确查找
+
+在用户名或者摘要中查找 `web user` 为关键字的用户。
+
+![image-20200715134811989](image/image-20200715134811989.png)
\ No newline at end of file
diff --git a/docs/zh-CN/src/vector.md b/docs/zh-CN/src/vector.md
new file mode 100644
index 0000000..3444bbb
--- /dev/null
+++ b/docs/zh-CN/src/vector.md
@@ -0,0 +1 @@
+# 向量那些事儿
diff --git a/src/pserver/simba/aggregation/function.rs b/src/pserver/simba/aggregation/function.rs
index 2279318..18bd73c 100644
--- a/src/pserver/simba/aggregation/function.rs
+++ b/src/pserver/simba/aggregation/function.rs
@@ -253,7 +253,7 @@ impl Hits {
     }
 
     pub fn map(&mut self, value: Vec<u8>) -> ASResult<bool> {
-        if self.result.hits.len() > self.result.size as usize {
+        if self.result.hits.len() >= self.result.size as usize {
             return Ok(false);
         }
         let result = &mut self.result;