使用 cstore fdw 加速hasura graphql-engine的查询

列式存储在数据分析（olap）上有天然的优势，以下是使用cstore fdw + hasura graphql-engine 加速
数据的分析

环境准备

• docker-compose 文件

 

version: '3'

services:

  graphql-engine:

    image: hasura/graphql-engine:v1.1.0-beta.3

    ports:

    - "8080:8080"

    environment:

      HASURA_GRAPHQL_DATABASE_URL: postgres://postgres:dalong@pgspider-cstore:5432/postgres

      HASURA_GRAPHQL_ENABLE_CONSOLE: "true" # set to "false" to disable console

      HASURA_GRAPHQL_ENABLED_LOG_TYPES: startup, http-log, webhook-log, websocket-log, query-log

  pgspider-pg_cron:

    image: dalongrong/pgspider:pg_cron

    ports: 

    - "5432:5432"

    environment: 

    - "POSTGRES_PASSWORD=dalong"

  pgspider-cstore:

    image: dalongrong/pgspider:cstore

    ports: 

    - "5433:5432"

    environment: 

    - "POSTGRES_PASSWORD=dalong"

• 启动&&数据准备

启动docker-compose 服务

docker-compose up -d

下载demo 数据

wget http://examples.citusdata.com/customer_reviews_1998.csv.gz

wget http://examples.citusdata.com/customer_reviews_1999.csv.gz

加压文件：

gzip -d customer_reviews_1998.csv.gz

gzip -d customer_reviews_1999.csv.gz

copy 数据到容器，使用docker cp 命令

使用cstore fdw

• 启用扩展

CREATE EXTENSION cstore_fdw;

• 创建server

CREATE SERVER cstore_server FOREIGN DATA WRAPPER cstore_fdw;

• 创建外部表

CREATE FOREIGN TABLE customer_reviews

(

    customer_id TEXT,

    review_date DATE,

    review_rating INTEGER,

    review_votes INTEGER,

    review_helpful_votes INTEGER,

    product_id CHAR(10),

    product_title TEXT,

    product_sales_rank BIGINT,

    product_group TEXT,

    product_category TEXT,

    product_subcategory TEXT,

    similar_product_ids CHAR(10)[]

)

SERVER cstore_server

OPTIONS(compression 'pglz');

• 导入数据

  容器内部

导入数据：

COPY customer_reviews FROM 'customer_reviews_1998.csv' WITH CSV;

COPY customer_reviews FROM 'customer_reviews_1999.csv' WITH CSV;

hasura graphql-engine 集成

• 添加track

• 数据查询

  此查询是通过graphql 的，实际和以前demo 的sql是一样的

参考资料

https://github.com/citusdata/cstore_fdw
https://github.com/citusdata/pg_cron
https://github.com/rongfengliang/pgspider-docker