docker-base | get process supervision and startup mechanisms | Continuous Deployment library

I finally managed to get it running by installing Java on the AirflowWorker container, as suggested by @floating_hammer.

See How to install java in an airflow container using docker-compose.yaml for details.

The encoding problem is solved using mysql-connector-python drivers, instead of the default mysqlclient!

• Append mysql-connector-python to ./docker/requirements-local.txt
• from superset, connect to mysql with this driver, using the following sqlalchemy uri mysql+mysqlconnector://:@[:]/

It's not really feasible:

• The end user would have to have the Docker runtime installed, which isn't a normal external dependency for a desktop application;
• The setup to run GUI applications in Docker is complicated and very platform-dependent (on non-Linux you need extra host-side software);
• Running Docker containers requires administrator-level permissions, since you can use a bind mount to edit arbitrary files on the host, but desktop applications don't usually have that.

(In principle you could work around all of these things, but Docker is intended to run as a system daemon and isn't designed to be embedded in other applications.)

The most straightforward packaging approach for this would be to publish the images to Docker Hub or another publicly-accessible repository, and to publish the docker-compose.yml file that launches the containers. Then an end user would need to install Docker, download the docker-compose.yml file, run docker-compose up -d, and point their browser to a specific URL to reach the front-end. That's less user-friendly than a standalone installer, but it is a small specific set of steps to run the application.

If you can restructure the application to use an embedded database like SQLite, and as a desktop application you don't need a federated login system, then you could also probably rebuild this whole thing as a native Electron application, without involving Docker at all.

I examined the architecture that you want to use multiple schemas (data sources, jdbc templates) in a single application. @Transactional only manages application's default data source that is named HikariPool-1. When you call the rest method, new hikari pool will be created that is named HikariPool-2. Your operations are on HikariPool-2, but @Transactional manages only HikariPool-1.

@Transactional's transaction manager argument can not be changed dynamically. So, you can define a new custom annotation that manages your transactions. Or you can use TransactionTemplate instead of annotations.

I created a simple custom transaction management aspect. It is working with defined dao's data source and transaction lifecycle.

just add the network_mode in YAML in db and geoserver and set it to host

A working solution is based on the David Maze and Matt's comments and on this question. These are the steps:

1. Open Docker for Mac, Preferences and Docker Engine. Add "ip": "127.0.0.1" to the json config file.
2. In the docker-compose, set ports to "127.0.0.1:5078:5078" for the web service.
3. Leave the Dockerfile and the python code as before: ie, the flask host is still 0.0.0.0

As I checked, messages from Electron's localhost 4201 went through. Also, running netstat -anvp tcp | awk 'NR<3 || /LISTEN/ shows that the insecure port 0.0.0.0.5078 is no longer exposed to the outside. `

Try sending a \r\n or \0 at the end of your message. The message is being sent as bytes over the network so it's probably being stored in buffers and the code reading the buffer needs a way to know the message is over. The regex is also matching on line terminators so will be necessary I think there as well.

Make definitely considered the second dependency. If it didn't "execute" then it means make decided nothing needed to be done to execute it.

I can see this:

While exploring the problem, I found out about recursive_error_pages on; -- but stupidly left it commented. Naturally it didn't help.

Un-commenting that directive fixes the issue!