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
- postgres
- kafka
- zookeeper
- redis
- 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
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 = false127.0.0.1 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.internalcore: 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
到此这篇关于docker内服务访问宿主机服务的实现的文章就介绍到这了,更多相关docker访问宿主机内容请搜索服务器之家以前的文章或继续浏览下面的相关文章希望大家以后多多支持服务器之家！
原文链接：https://www.cnblogs.com/failymao/p/15431490.html
