挂载卷问题：
docker-compose去做容器编排，可能会遇到volume挂载卷写错的问题，也就是从别的机器上复制下来的docker-compose.yml文件里面的volume是错的，这时候我们如果使用docker-compose up -d起动了，那么对应的volume会创建一个新的
比如我在别的服务器上对应docker-compose.yml文件volume配的是下面这样子：
volumes:
prometheus:
driver: local
driver_opts:
type: none
o: bind
device: /home/liushiwen555/prom/prometheus/data
grafana:
driver: local
driver_opts:
type: none
o: bind
device: /home/liushiwen555/prom/grafana
现在服务器换到别的地方后，原本的device路径要更换成/opt/prom/prometheus/data，在没修改的情况下，我们docker-compose up -d起了服务后，会报下面错误
Starting prom_dingtalk_1 ... done
Starting prom_node-exporter_1 ... done
Starting prom_alertmanager_1 ... done
Creating prom_prometheus_1 ... error
ERROR: for prom_prometheus_1 Cannot create container for service prometheus: error while mounting volume with options: type='none' device='/home/liushiwen555/prom/prometheus/data' o='bind': no such file or directory
ERROR: for prometheus Cannot create container for service prometheus: error while mounting volume with options: type='none' device='/home/liushiwen555/prom/prometheus/data' o='bind': no such file or directory
ERROR: Encountered errors while bringing up the project.
这是由于挂在卷错误，这时如果我们将原本prometheus挂载卷配置从/home/liushiwen555/prom/prometheus/data修改成/opt/prom/prometheus/data，再起还是会报同样的错误
这是由于docker-compose在判断你的VOLUME NAME存在的时候，是不会去创建新的挂载卷，会继续用以前的配置，这就导致它还是用的/home/liushiwen555/prom/prometheus/data这个配置
具体我们可以使用docker volume ls可以看到所有的挂载卷
[root@localhost prom]# docker volume ls
DRIVER VOLUME NAME
local 3dc40abe4f0b7cd22bfbef880187b28e2322bf450f69bb5adafc35aaf2cc6648
local 7a680be0a6843212cb8cd35d1fc6dfc05a7b581d67f8513a0be27aec9c8f7770
local 898b780f932cc81c8b542713adb08d2bf1426bd87d7e25ee52036091a5e792ab
local d9ae1a59bd4042006153c6e93c33360ae07529f135b6e61e21ee226adfdfb7cb
local metersphere_ms-conf
local metersphere_ms-data
local metersphere_ms-kafka-data
local metersphere_ms-logs
local metersphere_ms-mysql-data
local metersphere_ms-prometheus-conf
local metersphere_ms-prometheus-data
local metersphere_ms-zookeeper-data
local prom_grafana
local prom_prometheus
我们使用docker volume inspect prom_prometheus的时候会发现以下内容：
[root@localhost prom]# docker volume inspect prom_prometheus
[
{
"CreatedAt": "2021-11-04T13:55:41+08:00",
"Driver": "local",
"Labels": {
"com.docker.compose.project": "prom",
"com.docker.compose.volume": "prometheus"
},
"Mountpoint": "/var/lib/docker/volumes/prom_prometheus/_data",
"Name": "prom_prometheus",
"Options": {
"device": "/home/liushiwen555/prom/prometheus/data",
"o": "bind",
"type": "none"
},
"Scope": "local"
}
]
[root@localhost prom]# docker volume inspect prom_grafana
[
{
"CreatedAt": "2021-11-04T13:55:41+08:00",
"Driver": "local",
"Labels": {
"com.docker.compose.project": "prom",
"com.docker.compose.volume": "grafana"
},
"Mountpoint": "/var/lib/docker/volumes/prom_grafana/_data",
"Name": "prom_grafana",
"Options": {
"device": "/home/liushiwen555/prom/grafana",
"o": "bind",
"type": "none"
},
"Scope": "local"
}
]
会发现Options里面对应的device还是以前的 "device": "/home/liushiwen555/prom/prometheus/data"，我们新改的配置并没有应用
解决办法是得先暂停对应的容器实例：
docker stop 容器的CONTAINER ID
移除容器实例：
docker rm 容器的CONTAINER ID
删掉对应的挂载卷VOLUME NAME为prom_grafana和prom_prometheus
docker volume rm prom_grafana
docker volume rm prom_prometheus
[root@localhost prom]# docker volume ls
DRIVER VOLUME NAME
local 3dc40abe4f0b7cd22bfbef880187b28e2322bf450f69bb5adafc35aaf2cc6648
local 7a680be0a6843212cb8cd35d1fc6dfc05a7b581d67f8513a0be27aec9c8f7770
local 898b780f932cc81c8b542713adb08d2bf1426bd87d7e25ee52036091a5e792ab
local d9ae1a59bd4042006153c6e93c33360ae07529f135b6e61e21ee226adfdfb7cb
local metersphere_ms-conf
local metersphere_ms-data
local metersphere_ms-kafka-data
local metersphere_ms-logs
local metersphere_ms-mysql-data
local metersphere_ms-prometheus-conf
local metersphere_ms-prometheus-data
local metersphere_ms-zookeeper-data
local prom_grafana
local prom_prometheus
[root@localhost prom]# docker volume rm prom_grafana
prom_grafana
[root@localhost prom]# docker volume rm prom_prometheus
prom_prometheus
然后重新docker-compose up -d新的docker-compose.yml
最后对应的VOLUME会重新创建
[root@localhost prom]# docker-compose up -d
Creating volume "prom_prometheus" with local driver
Creating volume "prom_grafana" with local driver
prom_node-exporter_1 is up-to-date
prom_dingtalk_1 is up-to-date
prom_alertmanager_1 is up-to-date
Creating prom_prometheus_1 ... done
Creating prom_grafana_1 ... done
[root@localhost prom]# docker volume ls
DRIVER VOLUME NAME
local 3dc40abe4f0b7cd22bfbef880187b28e2322bf450f69bb5adafc35aaf2cc6648
local 7a680be0a6843212cb8cd35d1fc6dfc05a7b581d67f8513a0be27aec9c8f7770
local 898b780f932cc81c8b542713adb08d2bf1426bd87d7e25ee52036091a5e792ab
local d9ae1a59bd4042006153c6e93c33360ae07529f135b6e61e21ee226adfdfb7cb
local metersphere_ms-conf
local metersphere_ms-data
local metersphere_ms-kafka-data
local metersphere_ms-logs
local metersphere_ms-mysql-data
local metersphere_ms-prometheus-conf
local metersphere_ms-prometheus-data
local metersphere_ms-zookeeper-data
local prom_grafana
local prom_prometheus
[root@localhost prom]# docker volume inspect prom_grafana
[
{
"CreatedAt": "2021-11-04T15:14:13+08:00",
"Driver": "local",
"Labels": {
"com.docker.compose.project": "prom",
"com.docker.compose.volume": "grafana"
},
"Mountpoint": "/var/lib/docker/volumes/prom_grafana/_data",
"Name": "prom_grafana",
"Options": {
"device": "/opt/prom/grafana",
"o": "bind",
"type": "none"
},
"Scope": "local"
}
]
[root@localhost prom]# docker volume inspect prom_prometheus
[
{
"CreatedAt": "2021-11-04T15:14:11+08:00",
"Driver": "local",
"Labels": {
"com.docker.compose.project": "prom",
"com.docker.compose.volume": "prometheus"
},
"Mountpoint": "/var/lib/docker/volumes/prom_prometheus/_data",
"Name": "prom_prometheus",
"Options": {
"device": "/opt/prom/prometheus/data",
"o": "bind",
"type": "none"
},
"Scope": "local"
}
]
最后对应的服务就全起起来了端口映射问题：
docker里端口映射关系是 HOST:CONTAINER 的关系，遇到出现映射出来的HOST端口冲突，需要去修改端口映射
比如下面这种：
会发现端口映射写错了，左边宿主机端口，右边容器端口，prometheus默认端口是9090，而这里的端口映射关系写的9200:9200
右边端口应该是容器里prometheus的端口9090，这样写成9200那就需要去容器里改prometheus默认端口了，很麻烦，换成9200:9090就可以了，映射关系是HOST:CONTAINER 宿主机：容器
