docker-containers | Dockerfiles and assets for building Docker | Continuous Deployment library

You defined the postgres superuser as POSTGRES_USER: my_app (see the readme on dockerhub for details), but try to login with username postgres (psql -U **postgres**), which doese not exist. If we switch it to my_app, it should work. Same goes for the health check.

Alternatively, we could set the POSTGRES_USER to, e.g., postgres or root (or whatever we want) and provide initialization scripts to the container's directory /docker-entrypoint-initdb.d/. All *.sql-scripts in this folder will be executed on database startup. A common approach is to mount a host-directory, containing all scripts (e.g. creation of a new role, creation of a new database, and giving the new role ownership to this database) to this container-directory. An example can be found in this bitbucket repository of mine. We can just focus on the postgres container and the postgres directory, since those are the important parts.

You have the command already: /opt/kafka/bin/kc.sh.

I understand that you want to make it available to your users as a simple kc.sh or kc, without the directory prefix. The solution is to either put that directory into the PATH variable, typically done in /etc/profile; or put the command respectively a (symbolic) link to it into a directory which is already in the PATH, like /bin/, /usr/bin or /usr/local/bin.

Putting the directory into the PATH is probably the better solution, also because there are probably more useful commands to be found. If you don't like kc.sh but want a simple kc you can still make a link in the same directory, e.g. by performing cd /opt/kafka/binand then ln -s kc.sh kc.

I believe you have a problem, because AWS ECS https://github.com/aws/containers-roadmap/issues/938

Anyway, file system approach doesn't seems to be very cloud friendly (especially if you want to scale horizontally: problems with concurrent writes from multiple tasks, IOPs limitations, ...). Just provision proper DB (e.g. Aurora RDS Mysql, multi A-Z cluster if you need HA) and you will have nice opsless AWS deployment.

The safe way (and easier than with passwords too) is start using SSH keys and access containers over network. Or at least issue commands to host over SSH from gitlab-runner.

Also, SOF seach returned this: manage containers from another container, docker Looks legit.

You should change the RUN lines in the Dockerfile with apt-get in them to use the -y flag to skip asking you to confirm "yes".

The Dockerfile in the tutorial actually seems out of date or has errors in it. You need to use pip3 now to install Flask and also include the -y flag in your apt-get commands. I edited the Dockerfile from the tutorial and posted below:

Use the docker ps -l and -f flags for this. e.g.:

Docker can run on Windows 10 IoT Enterprise 64bit Build 16299 or later, with Hyper-V and Components enabled. See requirements.

Note: "Windows 10 IoT Enterprise" is identical to "Windows 10 Enterprise", it only differs in license type, not core system.

kdump.service is a system service, not a Docker container. You can read more on it here. What you see in logs is debug information, telling you that cAdvisor has no handler for kernel dump service. This is not an error and it is only visible because you've increased verbosity ("-v=4" in your command:). You can either decrease verbosity or simply ignore these messages.

Problem solved by Exiting the antivirus. even pausing it didn't helped because the firewall was still active.