Individual production docker script generator #974
Conversation
There was a problem hiding this comment.
Thanks @nedvedba for doing this, this greatly simplifies the process. Just a few comments and one big change needed. These runtime scripts need to replace what is in the CI pipelines so they are being tested as part of the CI process you will need to change the contents in .gitlab/end-to-end.yml.
I would also recommend putting comments in the auto generated files indicating that they were auto generated by the ./scripts/generate_ blahblahblah
| exit 1 | ||
| fi | ||
|
|
||
| cat << EOF > "$DATAFED_INSTALL_PATH/scripts/create_datafed_network.sh" |
There was a problem hiding this comment.
This will fail if the scripts folder is not first created you probably need to run
mkdir -p "${DATAFED_INSTALL_PATH}/scripts"
| -e DATAFED_DEFAULT_LOG_PATH="/datafed/logs" \\ | ||
| -e DATAFED_CORE_ADDRESS_PORT_INTERNAL="\$DATAFED_CORE_ADDRESS_PORT_INTERNAL" \\ | ||
| -e UID="\$USER_ID" \\ | ||
| --network datafed-network \\ |
There was a problem hiding this comment.
I'm pretty sure we are going to need to remove --network and assume that each container may very well be running on a different machine.
| cat << EOF > "$DATAFED_INSTALL_PATH/scripts/stop_web_container.sh" | ||
| #!/bin/bash | ||
|
|
||
| docker container stop datafed-web_$local_DOCKER_TAG |
There was a problem hiding this comment.
Why doesn't this have a rm line like the others?
This pull request serves to integrate the changes that add bash scripts that will generate the necessary scripts for running DataFed in production using individual scripts for starting, stopping, and removing the docker containers.