docker内服务访问宿主机服务的实现

目录

· 1. 场景

· 2. 解决

· 3. 总结

· 4. 参考

1. 场景

使用windows， wsl2 进行日常开发测试工作。 但是wsl2经常会遇到网络问题。比如今天在测试一个项目，核心功能是将postgres 的数据使用开源组件synch 同步到clickhouse 这个工作。

测试所需组件

1. postgres

2. kafka

3. zookeeper

4. redis

5. synch容器

最开始测试时，选择的方案是， 将上述五个服务使用 docker-compose 进行编排， network_modules使用hosts模式， 因为考虑到kafka的监听安全机制，这种网络模式，无需单独指定暴露端口。

docker-compose.yaml 文件如下

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88

version: "3" services:   postgres:     image: failymao/postgres:12.7     container_name: postgres     restart: unless-stopped     privileged: true                                                      # 设置docker-compose env 文件     command: [ "-c", "config_file=/var/lib/postgresql/postgresql.conf", "-c", "hba_file=/var/lib/postgresql/pg_hba.conf" ]     volumes:       - ./config/postgresql.conf:/var/lib/postgresql/postgresql.conf       - ./config/pg_hba.conf:/var/lib/postgresql/pg_hba.conf     environment:       POSTGRES_PASSWORD: abc123       POSTGRES_USER: postgres       POSTGRES_PORT: 15432       POSTGRES_HOST: 127.0.0.1     healthcheck:       test: sh -c "sleep 5 && PGPASSWORD=abc123 psql -h 127.0.0.1 -U postgres -p 15432 -c '\q';"       interval: 30s       timeout: 10s       retries: 3     network_mode: "host"   zookeeper:     image: failymao/zookeeper:1.4.0     container_name: zookeeper     restart: always     network_mode: "host"   kafka:     image: failymao/kafka:1.4.0     container_name: kafka     restart: always     depends_on:       - zookeeper     environment:       KAFKA_ADVERTISED_HOST_NAME: kafka       KAFKA_ZOOKEEPER_CONNECT: localhost:2181       KAFKA_LISTENERS: PLAINTEXT://127.0.0.1:9092       KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://127.0.0.1:9092       KAFKA_BROKER_ID: 1       KAFKA_LOG_RETENTION_HOURS: 24       KAFKA_LOG_DIRS: /data/kafka-data  #数据挂载     network_mode: "host"   producer:     depends_on:       - redis       - kafka       - zookeeper     image: long2ice/synch     container_name: producer     command: sh -c "       sleep 30 &&       synch --alias pg2ch_test produce"     volumes:       - ./synch.yaml:/synch/synch.yaml     network_mode: "host"   # 一个消费者消费一个数据库   consumer:     tty: true     depends_on:       - redis       - kafka       - zookeeper     image: long2ice/synch     container_name: consumer     command: sh -c       "sleep 30 &&       synch --alias pg2ch_test consume --schema pg2ch_test"     volumes:       - ./synch.yaml:/synch/synch.yaml     network_mode: "host"   redis:     hostname: redis     container_name: redis     image: redis:latest     volumes:       - redis:/data     network_mode: "host" volumes:   redis:   kafka:   zookeeper:

测试过程中因为要使用 postgres, wal2json组件,在容器里单独安装组件很麻烦， 尝试了几次均已失败而告终，所以后来选择了将 postgres 服务安装在宿主机上， 容器里面的synch服务 使用宿主机的 ip,port端口。

但是当重新启动服务后，synch服务一直启动不起来， 日志显示 postgres 无法连接. synch配置文件如下

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

core:   debug: true # when set True, will display sql information.   insert_num: 20000 # how many num to submit,recommend set 20000 when production   insert_interval: 60 # how many seconds to submit,recommend set 60 when production   # enable this will auto create database `synch` in ClickHouse and insert monitor data   monitoring: true redis:   host: redis   port: 6379   db: 0   password:   prefix: synch   sentinel: false # enable redis sentinel   sentinel_hosts: # redis sentinel hosts     - 127.0.0.1:5000   sentinel_master: master   queue_max_len: 200000 # stream max len, will delete redundant ones with FIFO source_dbs:   - db_type: postgres     alias: pg2ch_test     broker_type: kafka # current support redis and kafka     host: 127.0.0.1     port: 5433     user: postgres     password: abc123     databases:       - database: pg2ch_test         auto_create: true         tables:           - table: pgbench_accounts             auto_full_etl: true             clickhouse_engine: CollapsingMergeTree             sign_column: sign             version_column:             partition_by:             settings: clickhouse:   # shard hosts when cluster, will insert by random   hosts:     - 127.0.0.1:9000   user: default   password: ''   cluster_name:  # enable cluster mode when not empty, and hosts must be more than one if enable.   distributed_suffix: _all # distributed tables suffix, available in cluster kafka:   servers:     - 127.0.0.1:9092   topic_prefix: synch

这种情况很奇怪，首先确认 postgres, 启动，且监听端口(此处是5433) 也正常，使用localhost和主机eth0网卡地址均报错。

2. 解决

google 答案，参考 stackoverflow 高赞回答，问题解决，原答案如下

If you are using Docker-for-mac or Docker-for-Windows 18.03+, just connect to your mysql service using the host host.docker.internal (instead of the 127.0.0.1 in your connection string).

If you are using Docker-for-Linux 20.10.0+, you can also use the host host.docker.internal if you started your Docker

container with the --add-host host.docker.internal:host-gateway option.

Otherwise, read below

Use** --network="host" **in your docker run command, then 127.0.0.1 in your docker container will point to your docker host.

更多详情见 源贴

host 模式下 容器内服务访问宿主机服务

将postgres监听地址修改如下 host.docker.internal 报错解决。 查看宿主机 /etc/hosts 文件如下

1 2 3 4 5 6 7

root@failymao-NC:/mnt/d/pythonProject/pg_2_ch_demo# cat /etc/hosts # This file was automatically generated by WSL. To stop automatic generation of this file, add the following entry to /etc/wsl.conf: # [network] # generateHosts = false 127.0.0.1       localhost 10.111.130.24    host.docker.internal

可以看到，宿主机 ip跟域名的映射. 通过访问域名，解析到宿主机ip, 访问宿主机服务。

最终启动 synch 服务配置如下

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104

core:   debug: true # when set True, will display sql information.   insert_num: 20000 # how many num to submit,recommend set 20000 when production   insert_interval: 60 # how many seconds to submit,recommend set 60 when production   # enable this will auto create database `synch` in ClickHouse and insert monitor data   monitoring: true redis:   host: redis   port: 6379   db: 0   password:   prefix: synch   sentinel: false # enable redis sentinel   sentinel_hosts: # redis sentinel hosts     - 127.0.0.1:5000   sentinel_master: master   queue_max_len: 200000 # stream max len, will delete redundant ones with FIFO source_dbs:   - db_type: postgres     alias: pg2ch_test     broker_type: kafka # current support redis and kafka     host: host.docker.internal     port: 5433     user: postgres     password: abc123     databases:       - database: pg2ch_test         auto_create: true         tables:           - table: pgbench_accounts             auto_full_etl: true             clickhouse_engine: CollapsingMergeTree             sign_column: sign             version_column:             partition_by:             settings: clickhouse:   # shard hosts when cluster, will insert by random   hosts:     - 127.0.0.1:9000   user: default   password: ''   cluster_name:  # enable cluster mode when not empty, and hosts must be more than one if enable.   distributed_suffix: _all # distributed tables suffix, available in cluster kafka:   servers:     - 127.0.0.1:9092   topic_prefix: synch    host: host.docker.internal core:   debug: true # when set True, will display sql information.   insert_num: 20000 # how many num to submit,recommend set 20000 when production   insert_interval: 60 # how many seconds to submit,recommend set 60 when production   # enable this will auto create database `synch` in ClickHouse and insert monitor data   monitoring: true redis:   host: redis   port: 6379   db: 0   password:   prefix: synch   sentinel: false # enable redis sentinel   sentinel_hosts: # redis sentinel hosts     - 127.0.0.1:5000   sentinel_master: master   queue_max_len: 200000 # stream max len, will delete redundant ones with FIFO source_dbs:   - db_type: postgres     alias: pg2ch_test     broker_type: kafka # current support redis and kafka     host:     port: 5433     user: postgres     password: abc123     databases:       - database: pg2ch_test         auto_create: true         tables:           - table: pgbench_accounts             auto_full_etl: true             clickhouse_engine: CollapsingMergeTree             sign_column: sign             version_column:             partition_by:             settings: clickhouse:   # shard hosts when cluster, will insert by random   hosts:     - 127.0.0.1:9000   user: default   password: ''   cluster_name:  # enable cluster mode when not empty, and hosts must be more than one if enable.   distributed_suffix: _all # distributed tables suffix, available in cluster kafka:   servers:     - 127.0.0.1:9092   topic_prefix: synch

3. 总结

以--networks="host" 模式下启动容器时，如果想在容器内访问宿主机上的服务， 将ip修改为`host.docker.internal`

4. 参考

https:///questions/24319662/from-inside-of-a-docker-container-how-do-i-connect-to-the-localhost-of-the-mach