Public | Automated Build

Last pushed: 2 months ago
Short Description
Dlib-Torch container
Full Description

Identification Server

An automated build of the identification server is available on DockerHub: https://hub.docker.com/r/matbloch/uids/

Installation (Windows, Docker)

  1. Download and install Docker Toolbox (Win/OSX need Boot2Docker VM to run docker): Download Link
  2. Enable Portmapping from Host system to Boot2Docker VM:
    • Open VirtualBox Manager, Go to Boot2Docker VM Settings > Networking
    • For Adapter 1 (NAT) set port forwarding under advanced settings
    • Protocoll: TCP, Host IP: -, Client IP: -, Port: 8080 (Server Port)
  3. Start Boot2Docker: Head to Docker Toolbox installation directory and run start.sh with Git shell
  4. Pull server image:
    $ docker pull matbloch/uids
    
  5. Start server with forwarded port 8080:
    $ winpty docker run -p 8080:8080 -ti matbloch/uids //bin/bash
    

Demos/Tests

Share Local Installation with dockermachine

Allows you to quickly make adjustments to the server code on any host.

  1. Share Local Ressource to VM
    • Open VirtualBox Manager, Go to Boot2Docker VM Settings > Shared Folders
    • Add target folder under the name "/User". "/User" and some other (predefineds) folders automatically mapped to the docker VM
  2. Start docker image and share "/Users" under a custom name using the parameter -v:
    • winpty docker run -v //Users:/mycustomdir -ti matbloch/uids //bin/bash

test/tcp_connection Test TCP Connection: Host > boot2docker > Container

  1. Start docker container and expose port 8080
    winpty docker run -p 8080:8080 -ti matbloch/uids //bin/bash
    
  2. Start server on container (running on port 8080)
    python ../test/tcp_connection/server.py
    
  3. Start Python client on host machine
    python ../test/tcp_connection/client.py
    

Q/A and Problem Solving

Known Problems

1. Invalid Env. Variables

docker: An error occurred trying to connect: Post http://%2F%2F.%2Fpipe%2Fdocker
_engine/v1.24/containers/create: open //./pipe/docker_engine: Das System kann di
e angegebene Datei nicht finden..

Solution:

$ docker-machine env

Where C:\lib\Docker Toolbox\docker-machine.exe is the path to your docker machine:

$ eval $("C:\lib\Docker Toolbox\docker-machine.exe" env)

FAQ

My server is slow, how do I boost performance?

  • Assign more memory to the VM
  • Assign multiple cores to the VM (huge performance increase)
Docker Pull Command
Owner
matbloch
Source Repository

Comments (0)