Docker 安装 Citus 单节点集群：全面指南与详细操作

Docker 安装 Citus 单节点集群：全面指南与详细操作

本文详细介绍了如何使用 Docker 安装 Citus 单节点集群，并为其配置多个工作节点与协调器。通过具体的服务器资源表、部署示意图，以及 docker-compose 脚本，逐步指导读者如何创建 Docker 网络、配置 PostgreSQL 的 postgresql.conf 和 pg_hba.conf，并完成节点的添加与管理。此外，本文还提供了完整的 SQL 命令示例，帮助用户检查节点健康状况、查看节点表信息、创建分布式表和分片管理。无论是初学者还是有经验的开发者，都可以通过本文轻松完成 Citus 集群的安装与配置。

一 服务器资源

域名解释	服务器IP	端口	角色	备注
pg-cd1	192.168.0.1	15434	coordinator
pg-wk1	192.168.0.1	15432	worker
pg-wk2	192.168.0.2	15432	worker
pg-wk3	192.168.0.3	15432	worker

二 部署图

三 安装部署

1 创建网络

docker network create --driver bridge dbnet

2 运行脚本

1）docker-compose.cituscd1.yml

version: "3"services:pg-cd-1:image: citusdata/citus:12.1.3restart: alwayscontainer_name: pg-cd-1environment:- TZ=Asia/Shanghai- POSTGRES_USER=postgres- POSTGRES_PASSWORD=123456- PGUSER=postgres- PGPASSWORD=123456working_dir: /postgresqlports:- "15434:5432"networks:- dbnetextra_hosts:- pg-cd1:192.168.0.1- pg-wk1:192.168.0.1- pg-wk2:192.168.0.2- pg-wk3:192.168.0.3volumes:- ./postgresql/data:/var/lib/postgresql/data- ./postgresql/healthcheck-volume:/healthcheckhealthcheck:test: ["CMD-SHELL", "pg_isready -U postgres"]interval: 30stimeout: 10sretries: 5
networks:dbnet:external: true

2）docker-compose.cituswk1.yml

version: "3"services:pg-wk-1:image: citusdata/citus:12.1.3restart: alwayscontainer_name: pg-wk-1environment:- TZ=Asia/Shanghai- POSTGRES_USER=postgres- POSTGRES_PASSWORD=123456- PGUSER=postgres- PGPASSWORD=123456working_dir: /postgresqlnetworks:- dbnetextra_hosts:- pg-cd1:192.168.0.1- pg-wk1:192.168.0.1- pg-wk2:192.168.0.2- pg-wk3:192.168.0.3ports:- "15432:5432"volumes:- ./postgresql/data:/var/lib/postgresql/data- ./postgresql/healthcheck-volume:/healthcheckhealthcheck:test: ["CMD-SHELL", "pg_isready -U postgres"]interval: 30stimeout: 10sretries: 5
networks:dbnet:external: true

3）docker-compose.cituswk2.yml

version: "3"services:pg-wk-2:image: citusdata/citus:12.1.3restart: alwayscontainer_name: pg-wk-2environment:- TZ=Asia/Shanghai- POSTGRES_USER=postgres- POSTGRES_PASSWORD=123456- PGUSER=postgres- PGPASSWORD=123456working_dir: /postgresqlnetworks:- dbnetextra_hosts:- pg-cd1:192.168.0.1- pg-wk1:192.168.0.1- pg-wk2:192.168.0.2- pg-wk3:192.168.0.3ports:- "15432:5432"volumes:- ./postgresql/data:/var/lib/postgresql/data- ./postgresql/healthcheck-volume:/healthcheckhealthcheck:test: ["CMD-SHELL", "pg_isready -U postgres"]interval: 30stimeout: 10sretries: 5
networks:dbnet:external: true

4）docker-compose.cituswk3.yml

version: "3"services:pg-wk-3:image: citusdata/citus:12.1.3restart: alwayscontainer_name: pg-wk-3environment:- TZ=Asia/Shanghai- POSTGRES_USER=postgres- POSTGRES_PASSWORD=123456- PGUSER=postgres- PGPASSWORD=123456working_dir: /postgresqlnetworks:- dbnetextra_hosts:- pg-cd1:192.168.0.1- pg-wk1:192.168.0.1- pg-wk2:192.168.0.2- pg-wk3:192.168.0.3ports:- "15432:5432"volumes:- ./postgresql/data:/var/lib/postgresql/data- ./postgresql/healthcheck-volume:/healthcheckhealthcheck:test: ["CMD-SHELL", "pg_isready -U postgres"]interval: 30stimeout: 10sretries: 5
networks:dbnet:external: true

四 配置访问

1 修改 postgresql.conf 配置。

# Uncomment listen_addresses for the changes to take effect
listen_addresses = '*'

2 配置postgres客户端访问，修改 pg_hba.conf 。

# 每个节点都得配置，重启的时候有互相通信。
host    all             all             192.168.0.1/32        trust
host    all             all             192.168.0.3/32       trust
host    all             all             192.168.0.2/32       trust

五 节点添加

# 查看扩展
select * from pg_available_extensions;# 设置协调节点
SELECT citus_set_coordinator_host('pg-cd1', 15434);# 设置工作节点
SELECT citus_add_node('pg-wk1', 15432);
SELECT citus_add_node('pg-wk2', 15432);
SELECT citus_add_node('pg-wk3', 15432);# citus_get_active_worker_nodes() 函数返回活动工作线程主机名和端口号的列表。
SELECT * from citus_get_active_worker_nodes();# 检查所有节点之间的连通性
SELECT * FROM citus_check_cluster_node_health();# 集群中工作节点的信息
select * from pg_dist_node# 查看集群表信息
select * from citus_tables;# citus_total_relation_size 获取指定分布式表的所有分片使用的总磁盘空间
SELECT pg_size_pretty(citus_total_relation_size('devices'));# 查看表节点分布
SELECTdp.logicalrelid::regclass AS table_name,n.nodename,n.nodeport,pg_size_pretty(pg_total_relation_size(p.placementid::regclass)) AS shard_size
FROMpg_dist_partition dp
JOINpg_dist_shard s ON dp.logicalrelid = s.logicalrelid
JOINpg_dist_placement p ON s.shardid = p.shardid
JOINpg_dist_node n ON p.groupid = n.groupid
WHEREn.noderole = 'primary'
ORDER BYdp.logicalrelid,n.nodename;# 查看分片信息
select * from citus_shards

六 示例创建表

创建分布式表

CREATE TABLE events (id bigserial NOT NULL,device_id bigint,event_id bigserial,event_time timestamptz default now(),data jsonb not null,primary key (id)
);CREATE INDEX idx_dev_eve ON events (device_id, event_id);-- distribute the events table across shards placed locally or on the worker nodes
SELECT create_distributed_table('events', 'id', 'hash');INSERT INTO events (device_id, data)
SELECT s % 100, ('{"measurement":'||random()||'}')::jsonb FROM generate_series(1,10000) s;CREATE TABLE devices (device_id bigint primary key,device_name text,device_type_id int
);
CREATE INDEX ON devices (device_type_id);-- co-locate the devices table with the events table
SELECT create_distributed_table('devices', 'device_id', colocate_with := 'events');-- insert device metadata
INSERT INTO devices (device_id, device_name, device_type_id) 
SELECT s, 'device-'||s, 55 FROM generate_series(0, 99) s;