Passing from manual Docker host network to Docker Compose bridge

Question

I have 2 docker images a modbus server and a client which I run manually with docker run --network host server and the same with the client and work perfectly. But now I need to add them to a docker-compose file where the network is bridge, what I did like this:

autoserver:
    image: 19mikel95/pymodmikel:autoserversynchub
    container_name: autoserver
    restart: unless-stopped

  clientperf:
    image: 19mikel95/pymodmikel:reloadcomp
    container_name: clientperf
    restart: unless-stopped
    depends_on:
      - autoserver
    links:
      - "autoserver:server"

And I read that to refer from a container to another one(client to server) I have to use the service name in the dockercompose YML (autoserver) so that is what I did. In the python file executed in the client(which is the performance.py from pymodbus) I changed 'localhost' to:

host = 'autoserver'
client = ModbusTcpClient(host, port=5020)

However I get this error:

[ERROR/MainProcess] failed to run test successfully Traceback (most recent call last): File "performance.py", line 72, in single_client_test client.read_holding_registers(10, 1, unit=1) File "/usr/lib/python3/dist-packages/pymodbus/client/common.py", line 114, in read_holding_registers return self.execute(request) File "/usr/lib/python3/dist-packages/pymodbus/client/sync.py", line 107, in execute raise ConnectionException("Failed to connect[%s]" % (self.str())) pymodbus.exceptions.ConnectionException: Modbus Error: [Connection] Failed to connect[ModbusTcpClient(autoserver:5020)]

as asked, my full docker-compose YML is this:

version: '2.1'

networks:
  monitor-net:
    driver: bridge

volumes:
    prometheus_data: {}
    grafana_data: {}

services:

  prometheus:
    image: prom/prometheus:latest
    container_name: prometheus
    volumes:
      - ./prometheus:/etc/prometheus
      - prometheus_data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--web.console.libraries=/etc/prometheus/console_libraries'
      - '--web.console.templates=/etc/prometheus/consoles'
      - '--storage.tsdb.retention.time=200h'
      - '--web.enable-lifecycle'
    restart: unless-stopped
    expose:
      - 9090
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  alertmanager:
    image: prom/alertmanager:latest
    container_name: alertmanager
    volumes:
      - ./alertmanager:/etc/alertmanager
    command:
      - '--config.file=/etc/alertmanager/config.yml'
      - '--storage.path=/alertmanager'
    restart: unless-stopped
    expose:
      - 9093
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  nodeexporter:
    image: prom/node-exporter:latest
    container_name: nodeexporter
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - c:\:/rootfs:ro
    command:
      - '--path.procfs=/host/proc'
      - '--path.rootfs=/rootfs'
      - '--path.sysfs=/host/sys'
      - '--collector.filesystem.ignored-mount-points=^/(sys|proc|dev|host|etc)($$|/)'
    restart: unless-stopped
    expose:
      - 9100
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  cadvisor:
    image: gcr.io/google-containers/cadvisor:latest
    container_name: cadvisor
    volumes:
      - c:\:/rootfs:ro
      - /var/run:/var/run:rw
      - /sys:/sys:ro
      - /var/lib/docker:/var/lib/docker:ro
      #- /cgroup:/cgroup:ro #doesn't work on MacOS only for Linux
    restart: unless-stopped
    expose:
      - 8080
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  grafana:
    image: grafana/grafana:latest
    container_name: grafana
    volumes:
      - grafana_data:/var/lib/grafana
      - ./grafana/provisioning:/etc/grafana/provisioning
    environment:
      - GF_SECURITY_ADMIN_USER=${ADMIN_USER:-admin}
      - GF_SECURITY_ADMIN_PASSWORD=${ADMIN_PASSWORD:-admin}
      - GF_USERS_ALLOW_SIGN_UP=false
    restart: unless-stopped
    expose:
      - 3000
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  pushgateway:
    image: prom/pushgateway:latest
    container_name: pushgateway
    restart: unless-stopped
    expose:
      - 9091
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  caddy:
    image: stefanprodan/caddy
    container_name: caddy
    ports:
      - "3000:3000"
      - "9090:9090"
      - "9093:9093"
      - "9091:9091"
    volumes:
      - ./caddy:/etc/caddy
    environment:
      - ADMIN_USER=${ADMIN_USER:-admin}
      - ADMIN_PASSWORD=${ADMIN_PASSWORD:-admin}
    restart: unless-stopped
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  autoserver:
    image: 19mikel95/pymodmikel:autoserversynchub
    container_name: autoserver
    ports:
      - "5020:5020"
    restart: unless-stopped
    networks:
      - monitor-net

  clientperf:
    image: 19mikel95/pymodmikel:reloadcomp
    container_name: clientperf
    restart: unless-stopped
    networks:
      - monitor-net
    depends_on:
      - autoserver
    links:
      - "autoserver:server"

Have you verified that the `autoserver` container is running and ready to accept connections? — leopal, Jul 14 '20 at 08:50
@19mike95 could you please share the whole `docker-compose.yaml` file? — Dmytro Serdiuk, Jul 14 '20 at 08:52
Did you try creating an env variable for host and give that as the host on your code? — Subbu, Jul 14 '20 at 08:55
@DmytroSerdiuk done, although I have done some changes and it might be not be exactly the same of the part I posted the first time — 19mike95, Jul 14 '20 at 08:57
@19mike95 I see `host= 'autoserver'` on the code that you provided. Here python considering 'autoserver' as string — Subbu, Jul 14 '20 at 09:00
@Subbu yes, however i think that is how its supposed to work, because if llok into the performance.py file from pymodbus they use host = '127.0.0.1' to specify localhost. — 19mike95, Jul 14 '20 at 09:03
I think you don't need the `links` in `clientperf`. When using the same networks, it can be accessible via host name already. — Thinh Tran, Jul 14 '20 at 09:09
@ThinhTran yes i know it is not needed, however as it was not working I tried just in case — 19mike95, Jul 14 '20 at 09:27
@19mike95 It's strange. Have you tried to run some dns troubleshooting commands like `nslookup autoserver` or `dig autoserver` on your client container? — Thinh Tran, Jul 14 '20 at 09:43

score 1 · Accepted Answer · answered Jul 14 '20 at 10:40

The problem is in the StartTcpServer(context, identity=identity, address=("localhost", 5020)) of sincserver.py file from autoserver image. The localhost allows for TcpServer to accept connection only from localhost. It should be replaced with 0.0.0.0 in order to allow any external requests to this port.

The following Docker Compose shows it (sed -i 's|localhost|0.0.0.0|g' sincserver.py replaces the hostname):

version: '2.1'

services:
  autoserver:
    image: 19mikel95/pymodmikel:autoserversynchub
    command: sh -c "
      sed -i 's|localhost|0.0.0.0|g' sincserver.py;
      python3 sincserver.py daemon off
      "
    ports:
      - "5020:5020"
    restart: unless-stopped

  clientperf:
    image: 19mikel95/pymodmikel:reloadcomp
    restart: unless-stopped
    depends_on:
      - autoserver

Run:

docker-compose up -d
docker-compose logs -f clientperf

And you will see log like

clientperf_1  | [DEBUG/MainProcess] 574 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.7410697999875993 seconds
clientperf_1  | [DEBUG/MainProcess] 692 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.4434449000109453 seconds
clientperf_1  | [DEBUG/MainProcess] 708 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.4116760999895632 seconds
clientperf_1  | [DEBUG/MainProcess] 890 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.1230684999900404 seconds
clientperf_1  | [DEBUG/MainProcess] 803 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.2450218999874778 seconds
clientperf_1  | [DEBUG/MainProcess] 753 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.3274328999978025 seconds
clientperf_1  | [DEBUG/MainProcess] 609 requests/second
clientperf_1  | [DEBUG/MainProcess] time taken to complete 1000 cycle by 10 workers is 1.6399398999928962 seconds

Subbu · Answer 2 · 2020-07-14T09:35:30.257

You can try using the env variable like AUTO_SERVER_HOST and call this in your code

version: '2.1'

networks:
  monitor-net:
    driver: bridge

volumes:
    prometheus_data: {}
    grafana_data: {}

services:

  prometheus:
    image: prom/prometheus:latest
    container_name: prometheus
    volumes:
      - ./prometheus:/etc/prometheus
      - prometheus_data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--web.console.libraries=/etc/prometheus/console_libraries'
      - '--web.console.templates=/etc/prometheus/consoles'
      - '--storage.tsdb.retention.time=200h'
      - '--web.enable-lifecycle'
    restart: unless-stopped
    expose:
      - 9090
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  alertmanager:
    image: prom/alertmanager:latest
    container_name: alertmanager
    volumes:
      - ./alertmanager:/etc/alertmanager
    command:
      - '--config.file=/etc/alertmanager/config.yml'
      - '--storage.path=/alertmanager'
    restart: unless-stopped
    expose:
      - 9093
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  nodeexporter:
    image: prom/node-exporter:latest
    container_name: nodeexporter
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - c:\:/rootfs:ro
    command:
      - '--path.procfs=/host/proc'
      - '--path.rootfs=/rootfs'
      - '--path.sysfs=/host/sys'
      - '--collector.filesystem.ignored-mount-points=^/(sys|proc|dev|host|etc)($$|/)'
    restart: unless-stopped
    expose:
      - 9100
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  cadvisor:
    image: gcr.io/google-containers/cadvisor:latest
    container_name: cadvisor
    volumes:
      - c:\:/rootfs:ro
      - /var/run:/var/run:rw
      - /sys:/sys:ro
      - /var/lib/docker:/var/lib/docker:ro
      #- /cgroup:/cgroup:ro #doesn't work on MacOS only for Linux
    restart: unless-stopped
    expose:
      - 8080
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  grafana:
    image: grafana/grafana:latest
    container_name: grafana
    volumes:
      - grafana_data:/var/lib/grafana
      - ./grafana/provisioning:/etc/grafana/provisioning
    environment:
      - GF_SECURITY_ADMIN_USER=${ADMIN_USER:-admin}
      - GF_SECURITY_ADMIN_PASSWORD=${ADMIN_PASSWORD:-admin}
      - GF_USERS_ALLOW_SIGN_UP=false
    restart: unless-stopped
    expose:
      - 3000
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  pushgateway:
    image: prom/pushgateway:latest
    container_name: pushgateway
    restart: unless-stopped
    expose:
      - 9091
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  caddy:
    image: stefanprodan/caddy
    container_name: caddy
    ports:
      - "3000:3000"
      - "9090:9090"
      - "9093:9093"
      - "9091:9091"
    volumes:
      - ./caddy:/etc/caddy
    environment:
      - ADMIN_USER=${ADMIN_USER:-admin}
      - ADMIN_PASSWORD=${ADMIN_PASSWORD:-admin}
    restart: unless-stopped
    networks:
      - monitor-net
    labels:
      org.label-schema.group: "monitoring"

  autoserver:
    image: 19mikel95/pymodmikel:autoserversynchub
    container_name: autoserver
    ports:
      - "5020:5020"
    restart: unless-stopped
    networks:
      - monitor-net

  clientperf:
    image: 19mikel95/pymodmikel:reloadcomp
    container_name: clientperf
    restart: unless-stopped
    networks:
      - monitor-net
    depends_on:
      - autoserver
    environment:
      - AUTO_SERVER_HOST=autoserver

Call env variable like below

host = os.environ['AUTO_SERVER_HOST']
client = ModbusTcpClient(host, port=5020)

`ERROR: The Compose file '.\docker-compose.yml' is invalid because: Unsupported config option for services.clientperf: 'env' ` and also tried with 'enviroment' which is the one supposed to work in docker-compose 2.1 but still got the same error. — 19mike95, Jul 14 '20 at 09:26
is there any specific reason for using the docker-compose version: '2.1'? — Subbu, Jul 14 '20 at 09:28
I am using drockprom and there is no newer version for docker-compose 3 — 19mike95, Jul 14 '20 at 09:29
managed to use the environment still got the same error as alwways `File "/usr/lib/python3/dist-packages/pymodbus/client/sync.py", line 107, in execute raise ConnectionException("Failed to connect[%s]" % (self.__str__())) pymodbus.exceptions.ConnectionException: Modbus Error: [Connection] Failed to connect[ModbusTcpClient(autoserver:5020)]` — 19mike95, Jul 14 '20 at 09:38
Let's debug it. can you print `os.environ['AUTO_SERVER_HOST']` and see what it is giving? — Subbu, Jul 14 '20 at 09:42
It is not needed it's returning 'autoserver' it can be seen in the error above ModbusTcpClient(**autoserver**:5020) — 19mike95, Jul 14 '20 at 09:44
Let us [continue this discussion in chat](https://chat.stackoverflow.com/rooms/217788/discussion-between-19mike95-and-subbu). — 19mike95, Jul 14 '20 at 09:54

Passing from manual Docker host network to Docker Compose bridge

2 Answers2