pipelinedb continuous view 操作

continuous view 是 pipelinedb的核心，类似一个view，但是数据是合并了stream以及table的数据输入数据，并且是
实时根据输入数据进行更新的

语法

CREATE CONTINUOUS VIEW name AS query

query是一个pg 的select 格式的语法，格式如下：

SELECT [ DISTINCT [ ON ( expression [, ...] ) ] ]
    expression [ [ AS ] output_name ] [, ...]
    [ FROM from_item [, ...] ]
    [ WHERE condition ]
    [ GROUP BY expression [, ...] ]
    [ WINDOW window_name AS ( window_definition ) [, ...] ]

where from_item can be one of:

    stream_name [ [ AS ] alias [ ( column_alias [, ...] ) ] ]
    table_name [ [ AS ] alias [ ( column_alias [, ...] ) ] ]
    from_item [ NATURAL ] join_type from_item [ ON join_condition ]

环境准备

项目使用docker运行同时结合hasura graphql 引擎

• docker-compose

version: '3.6'
services:
  postgres:
    image:  pipelinedb/pipelinedb
    ports:
    - "5432:5432"
  graphql-engine:
    image: hasura/graphql-engine:v1.0.0-alpha06
    ports:
    - "8080:8080"
    depends_on:
    - "postgres"
    command: >
      /bin/sh -c "
      graphql-engine --database-url postgres://pipeline:pipeline@postgres:5432/pipeline serve --enable-console;
      "

数据源从基本数据tabley以及stream 获取（比较综合的例子）

• 创建基本表

CREATE TABLE userlogin (
    id SERIAL PRIMARY KEY,
    username text NOT NULL,
    userid integer NOT NULL,
    usertype text NOT NULL,
    logintype text NOT NULL
);

 ```
* 创建stream：
```code
CREATE STREAM loginlogs (logintype text, userid integer);

• 创建continuous view

CREATE CONTINUOUS VIEW userloginview AS 

select a.logintype,b.username,b.userid, b.logintype as logintype_ from loginlogs a join  userlogin b

on a.userid=b.userid

• 插入数据&&查询
  

insert into  loginlogs(logintype,userid) values ('mobile',333),('pc',333),('web',333)

select * from userloginview

• graphql 集成
  
• graphql 查询
  
• 说明
  实际使用中我们的view一般都是一个聚合函数的操作，比如统计状态，异常信息排查,同时view 支持ttl 可以支持有效期控制

官方提供的一个比较有意思的demo

• 延迟百分比 90 95 99 延迟占比

CREATE CONTINUOUS VIEW latency AS
SELECT percentile_cont(array[90, 95, 99]) WITHIN GROUP (ORDER BY latency)
FROM latency_stream;

• 最新5分钟广告的曝光

CREATE CONTINUOUS VIEW imps AS
SELECT COUNT(*) FROM imps_stream
WHERE (arrival_timestamp > clock_timestamp() - interval '5 minutes');

参考资料

http://docs.pipelinedb.com/continuous-views.html