前面介绍了nginx与php两个容器间是如何进行通信的:
[root@docker ~]# docker run -d --name=php -v /www:/usr/local/nginx/html php
[root@docker ~]# docker run -d --name=nginx --link=php:php -v /www:/usr/local/nginx/html -p 81:80 nginx
[root@docker ~]# docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES 58280fe851f9 nginx "/usr/local/nginx/..." 15 seconds ago Up 14 seconds 0.0.0.0:81->80/tcp nginx 9ea150c35587 php "/usr/local/php/sb..." 36 seconds ago Up 35 seconds 9000/tcp php
通过容器间的选项--link指定容器名称进行不同容器间的通信(--link container_name或者将container_name取一个别名)
现在使用另外一种方式替代--link来达到容器间的通信:docker network
查看local的网络信息:
[root@docker ~]# docker network ls NETWORK ID NAME DRIVER SCOPE 5133ec415c3c bridge bridge local f359ca4e2d39 host host local 8d68673c045c none null local
现在创建一个网络名为my_net且driver为bridge的网络:(默认创建的就是bridge)
[root@docker ~]# docker network create my_net 67e29f0e4a77c79144efc337a081a889188b5b8e289968f22be6e4ddd9b80610 [root@docker ~]# docker network ls NETWORK ID NAME DRIVER SCOPE 5133ec415c3c bridge bridge local f359ca4e2d39 host host local 67e29f0e4a77 my_net bridge local 8d68673c045c none null local
利用--network启动容器提供服务:
[root@docker ~]# docker run -d --name=php --network my_net --network-alias php -v /www:/usr/local/nginx/html php 6b493cbe8207dee4cb4d5945cfce305dba96914083bd7f46841b0b42376bcb99 [root@docker ~]# docker run -d --name=nginx --network my_net --network-alias nginx -v /www:/usr/local/nginx/html -p 80:80 nginx 5ab220196b52bb768bef508433f0b920eecee70c3ee47880ebc5e2a74b5ee254
通过选项--network-alias将取名的my_net起了一个别名
[root@docker ~]# docker exec -it nginx ping php PING php (172.18.0.2) 56(84) bytes of data. 64 bytes from php.my_net (172.18.0.2): icmp_seq=1 ttl=64 time=0.079 ms 64 bytes from php.my_net (172.18.0.2): icmp_seq=2 ttl=64 time=0.090 ms
这里为啥能够ping通过php这个容器呢,因为这两个容器在同一个网络m'y_net内,而nginx里面ping的php是容器名(这里管理的是容器级别)
由于能够ping通php,所以在nginx中配置的:
server { listen 80; root /usr/local/nginx/html; index index.htm index.html index.php; location ~ .php$ { root /usr/local/nginx/html; fastcgi_pass php:9000; fastcgi_index index.php; fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; include fastcgi_params; } }
所以配置文件中的php不会导致nginx启动失败,通过network的driver bridge实现了容器间的访问
上面在容器启动的时候使用的是选项--network,而在compose的配置文件中则是networks,现在通过配置文件来进行阐述该参数的作用:
[root@docker lnmp]# cat lnmp.yml version: '3' services: nginx: image: nginx container_name: lnmp-nginx depends_on: - php ports: - "80:80" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" php: image: php container_name: lnmp-php expose: - "9000" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" networks: net1: driver: bridge
由上述文件中可以知道networks定义了一个名称为net1的网络,由于networks是top-level(顶层级别,所以需要在顶层设置),而在创建的网络时候需要指定driver(单一网络使用bridge,swarm集群使用overlay),而且driver内容不能省略,在nginx和php两个service中使用了同一网络net1,现在将服务启动:
[root@docker lnmp]# docker-compose -f lnmp.yml up Creating network "lnmp_net1" with driver "bridge" Creating lnmp-php ... Creating lnmp-php ... done Creating lnmp-nginx ... Creating lnmp-nginx ... done Attaching to lnmp-php, lnmp-nginx
可以看出在启动服务的时候创建了服务级别的网络lnmp_net1
[root@docker lnmp]# docker network ls NETWORK ID NAME DRIVER SCOPE 5133ec415c3c bridge bridge local f359ca4e2d39 host host local 29d798852b52 lnmp_net1 bridge local 67e29f0e4a77 my_net bridge local 8d68673c045c none null local
服务nginx和php在网络lnmp_net1实现了互联通信
[root@docker lnmp]# docker-compose -f lnmp.yml exec nginx ping php PING php (172.19.0.2) 56(84) bytes of data. 64 bytes from lnmp-php.lnmp_net1 (172.19.0.2): icmp_seq=1 ttl=64 time=0.060 ms
在nginx服务中能够访问php的服务,在compose编排中级别的访问是对service级别的访问,所以在nginx服务配置文件中对应的php能够实现,不会导致nginx启动失败
及在compose的配置文件设置并对应的是service级别,当使用参数links也是配置的php服务,使得nginx服务能够ping通php服务并实现访问通信
[root@docker lnmp]# cat lnmp.yml version: '3' services: nginx: image: nginx container_name: lnmp-nginx depends_on: - php ports: - "80:80" links: - php:php volumes: - "/www:/usr/local/nginx/html" php: image: php container_name: lnmp-php expose: - "9000" volumes: - "/www:/usr/local/nginx/html"
将networks参数改为了links参数,使得两个服务能够继续互联访问
[root@docker lnmp]# docker-compose -f lnmp.yml up Creating network "lnmp_default" with the default driver Creating lnmp-php ... Creating lnmp-php ... done Creating lnmp-nginx ... Creating lnmp-nginx ... done Attaching to lnmp-php, lnmp-nginx
而且启动的时候创建的默认的网络
[root@docker lnmp]# docker-compose -f lnmp.yml exec nginx ping php PING php (172.19.0.2) 56(84) bytes of data. 64 bytes from lnmp-php.lnmp_default (172.19.0.2): icmp_seq=1 ttl=64 time=0.091 ms 64 bytes from lnmp-php.lnmp_default (172.19.0.2): icmp_seq=2 ttl=64 time=0.107 ms
--links也是能够使得服务间进行访问的
现在通过案例讲解参数external_links的作用:external_links类似于links针对的也是service级别,但是external_links对应的是外部的service而不是同一compose配置文件中的service
[root@docker lnmp]# cat lnmp.yml version: '3' services: nginx: image: nginx container_name: lnmp-nginx depends_on: - php ports: - "80:80" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" external_links: - php1:php php: image: php container_name: lnmp-php expose: - "9000" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" networks: net1: driver: bridge
上面可以看见external_links配置了php1的服务,并且将php1的服务也取别名为了php,然后编写另一个compose配置文件,并配置服务名为php1的服务:
[root@docker lnmp]# cat php1.yml version: '3' services: php1: image: php container_name: lnmp-php1 expose: - "9000" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" networks: net1: driver: bridge
结合两个compose配置文件,可以知道有nginx、php、php1的服务,并且php1的服务也有php的别名,现在依次开启php1,和nginx与php:
[root@docker lnmp]# docker-compose -f php1.yml up Creating network "lnmp_net1" with driver "bridge" Creating lnmp-php1 ... Creating lnmp-php1 ... done Attaching to lnmp-php1
[root@docker lnmp]# docker-compose -f lnmp.yml up --build WARNING: Found orphan containers (lnmp-php1) for this project. If you removed or renamed this service in your compose file, you can run this command with the --remove-orphans flag to clean it up. Creating lnmp-php ... Creating lnmp-php ... done Creating lnmp-nginx ... Creating lnmp-nginx ... done Attaching to lnmp-php, lnmp-nginx
查看启动的容器状态:
[root@docker www]# docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES a089cd261279 nginx "/usr/local/nginx/..." 25 seconds ago Up 25 seconds 0.0.0.0:80->80/tcp lnmp-nginx 4ff4bf7c1bf9 php "/usr/local/php/sb..." 25 seconds ago Up 25 seconds 9000/tcp lnmp-php 4e39cdf62bf1 php "/usr/local/php/sb..." 51 seconds ago Up 51 seconds 9000/tcp lnmp-php1
现在将nginx一起的php的这个服务停止掉,看下nginx服务是否将会被停止:
[root@docker lnmp]# docker-compose -f lnmp.yml stop php Stopping lnmp-php ... done
而在nginx这里的输出如下:
lnmp-php exited with code 0
[root@docker www]# docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES a089cd261279 nginx "/usr/local/nginx/..." 2 minutes ago Up 2 minutes 0.0.0.0:80->80/tcp lnmp-nginx 4ff4bf7c1bf9 php "/usr/local/php/sb..." 2 minutes ago Exited (0) 52 seconds ago lnmp-php
而php服务对应的容器lnmp-php也停止了,但是nginx的服务却没有停止,为什么呢?那是因为php1的服务并没有停止,而php1的服务也是别名php服务的,而external_links就是实现这种情况的
使一个compose中的服务能与另一个compose中的服务能够互联通信,compose就是service级别的实现,配置文件中对应的都是service级别,而不是容器名
而在另一个compose中也必须定义与之相同的网络名称,也就是说配置external_links时,两者服务间的网络名称必须一致,不然依旧不能进行互相访问
当使用了external_links时,必须启动另一个compose的service,如果仅仅启动的是包含此函数的compose那么service将会启动报错
[root@docker lnmp]# cat lnmp.yml_external_links version: '3' services: nginx: image: nginx container_name: lnmp-nginx depends_on: - php ports: - "80:80" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" external_links: - php1:php php: image: php container_name: lnmp-php expose: - "9000" networks: - "net1" volumes: - "/www:/usr/local/nginx/html" networks: net1: driver: bridge
[root@docker lnmp]# docker-compose -f lnmp.yml up --build Creating network "lnmp_net1" with driver "bridge" Creating lnmp-php ... Creating lnmp-php ... done Creating lnmp-nginx ... Creating lnmp-nginx ... done Attaching to lnmp-php, lnmp-nginx lnmp-nginx | nginx: [emerg] host not found in upstream "php" in /usr/local/nginx/conf/vhost/www.conf:7 lnmp-nginx exited with code 1
参数external_links连接的外部php服务并没有启动,所以nginx还是启动报错