Collecting Society Docker Setup¶

Docker development and deployment setup for collecting society services.

The main resources can be found here:

Overview¶

The setup creates and maintains the docker services for development, deployment, testing, and documentation. The tool docker-compose is used as a creator and configurator for the docker services.

Schema¶

                                                       _
                                        ------------    |
       webbrowser           tryton      |  worker  |    | Clients
           .                  .         ------------   _|
           |                  |           |      |     _
-----------------------------------       |      |      |
|             webserver           |       |      |      | Public
-----------------------------------       |      |     _|
     |           |            |           |      |     _
----------   ----------   ---------------------  |      |
| webgui |   | webapi |   |      erpserver    |  |      |
----------   ----------   ---------------------  |      |
     |           |            |           |      |      | Internal
-----------------------------------  ---------------    |
|             database            |  | fingerprint |    |
-----------------------------------  ---------------   _|

Services¶

Service	Description	Repositories	Ports	Volumes
database	Postgres DB			shared postgresql-data
erpserser	Trytond Server	collecting_society	8000: jsonrpc 52005: debugpy 52006: debugpy	shared trytond-files
webserver	Nginx Server		80: http	shared nginx-certs nginx-dhparam nginx-htpasswd
webgui	Pyramid Gui App +Trytond Server	portal_web collecting_society_web	6543: pserve 52000: debugpy	shared trytond-files
webapi	Pyramid Api App +Trytond Server	portal_web collecting_society_web	6544: pserve 52001: debugpy	shared trytond-files
worker	File Processing +Proteus Client	collecting_society_worker	52002: debugpy	shared
fingerprint	Echoprint Server	echoprint-server	8080: http 52004: debugpy	shared echoprint-data

Files¶

Note

Some files and folders are created on the first run of the project script update command, docs-build script or service-test script and many are symlinks into the shared volume, as the files need to be accessible from within the container.

├── code/                               # service application repositories
│   ├── collecting_society/             # (erpserver) tryton module
│   ├── portal_web/                     # (webgui/webapi) pyramid main app
│   ├── collecting_society_web/         # (webgui/webapi) pyramid plugin app
│   ├── collecting_society_worker/      # (worker) file processing
│   └── echoprint-server/               # (fingerprint) echoprint server
│
├── docs/                               # documentation build
│   └── index.html                      # main index file of the built documentation
│
├── services/                           # files for docker services
│   ├── build/                          # build environment for docker images
│   │   ├── Dockerfile                  # multistage Dockerfile for docker images
│   │   ├── pip_pin.sh                  # pins pip versions in Dockerfile
│   │   ├── pip_unpin.sh                # unpins pip versions in Dockerfile
│   │   └── worker.cron                 # (worker) cronjob file for processing
│   │
│   ├── config/                         # config files for services
│   │   ├── collecting_society.*        # (erpserver) trytond config / passfile
│   │   ├── portal_web.*                # (webapi/gui) pyramid config
│   │   ├── collecing_society_web.*     # (webapi/gui) pyramid plugin config
│   │   ├── collecting_society_worker.* # (worker) worker config
│   │   └── nginx.proxy.conf            # (webserver) additional proxy wide nginx config
│   │
│   ├── deploy/                         # deployment scripts for services
│   ├── healthcheck/                    # healthcheck scripts for services
│   ├── pip/                            # pip runtime requirements for services
│   │
│   └── <SERVICE>.env                   # additional envvars file for SERVICE
│
├── tests/                              # testing output
│   ├── junit_erpserver.xml             # (erperserver) junit results
│   ├── cover_erpserver.xml             # (erperserver) coverage results (xml)
│   ├── cover_erpserver.html            # (erperserver) coverage results (html)
│   │   └── index.html                  # main index file of coverage results
│   ├── junit_web.xml                   # (webgui/webapi) junit results
│   ├── cover_web.xml                   # (webgui/webapi) coverage results (xml)
│   ├── cover_web.html/                 # (webgui/webapi) coverage results (html)
│   │   └── index.html                  # main index file of coverage results
│   ├── junit_worker.xml                # (worker) junit results
│   ├── cover_worker.xml                # (worker) coverage results (xml)
│   ├── cover_worker.html/              # (worker) coverage results (html)
│   │   └── index.html                  # main index file of coverage results
│   └── screenshots/                    # (webgui) screenshots of integration tests
│
├── volumes/                            # volumes mounted into the containers
│   ├── shared/                         # (*) main volume mounted into all containers
│   │   ├── src/                        # repos of packages to include on runtime
│   │   ├── ref/                        # repos of packages in docker images for reference
│   │   │
│   │   ├── data/                       # demodata generation module
│   │   │   ├── csv/                    # csv files to import
│   │   │   │   ├── <MODEL>.csv         # csv file for tryton MODEL
│   │   │   │   └── <MODEL>.py          # script to generate the csv file for tryton MODEL
│   │   │   ├── datasets/               # datasets to generate
│   │   │   │   └── <MODEL>.py          # dataset for tryton MODEL
│   │   │   ├── fingerprints/           # fingerprints for echoprint
│   │   │   ├── uploads/                # audiofile generation and compression script
│   │   │   ├── modelinfo.py            # outputs csv with model tryton ids and class names
│   │   │   └── main.py                 # main demodata generation script
│   │   │
│   │   ├── docs/                       # documentation sphinx build environment
│   │   │   ├── build/                  # build of the documentation
│   │   │   ├── source/                 # source of the documentation
│   │   │   ├── build.sh                # sphinx build script (run in container!)
│   │   │   └── Makefile                # sphinx Makefile
│   │   │
│   │   ├── tmp/                        # tmp data of services (development/testing)
│   │   │   ├── files/                  # trytond file storage
│   │   │   ├── logs/                   # log files for debugging
│   │   │   ├── sessions/               # cookie session files
│   │   │   └── upload/                 # file upload processing
│   │   │       └── <STAGE>/            # processing / archiving STAGE of files
│   │   │
│   │   ├── docker-entrypoint.sh        # docker entrypoint for python based containers
│   │   └── cli                         # main CLI script for common tasks (run in container!)
│   │
│   ├── echoprint-data/                 # (fingerprint) echoprint database data
│   ├── nginx-certs/                    # (webserver) certificates
│   ├── nginx-dhparam/                  # (webserver) dh parameters
│   ├── nginx-htpasswd/                 # (webserver) htpassword injection
│   ├── postgresql-data/                # (database) postgres database data
│   └── tryton-files/                   # (erpserver/webgui/webapi) trytond file storage
│
├── .env                                # main environment variable file
├── project.yml                         # main project setup file
│
├── project                             # updates the files/folders/repos of the project
├── db-rebuild                          # rebuilds the database
├── docs-build                          # builds the documentation of the project
├── service-test                        # runs the tests of the project
├── cli                                 # main CLI script for common tasks (run in container!)
│
├── docker-compose.yml                  # main docker compose file
├── docker-compose.override.yml         # symlink to environment docker override file
├── docker-compose.development.yml      # -> docker override file for development
├── docker-compose.staging.yml          # -> docker override file for staging
├── docker-compose.production.yml       # -> docker override file for production
├── docker-compose.testing.yml          # standalone docker compose file for testing
├── docker-compose.documentation.yml    # standalone docker compose file for documentation
│
├── .vscode/                            # settings for vs code
├── .devcontainer.json*                 # settings for vs code remote containers
├── .flake8                             # symlink to settings for flake8 linter
├── .gitignore                          # ignore patterns for git
├── .rgignore                           # ignore patterns for ripgrep
│
├── CHANGELOG.rst                       # changelog
├── COPYRIGHT.rst                       # copyright
├── LICENSE-AGPLv3.txt                  # license
└── README.rst                          # this readme

Docker¶

`.env`	Main .env environment file for service configuration
`docker-compose.yml`	Main docker compose file with the definition of the services
`docker-compose.override.yml`	Environments variables overriding those of the main file
`services/build/Dockerfile`	Multistage Dockerfile for the docker images
`volumes/shared/docker-entrypoint.sh`	Entrypoint script for python based containers

Development¶

`project.yml`	project setup configuration file for file/folder/repo tasks
`project`	project script for project maintainance tasks
`project update`	updates the files/folders/repos of the project
`project status`	status of all project repositories
`project diff`	diff of all project repositories and example files
`project pull`	pull all project repositories
`project checkout BRANCH`	checkout BRANCH in all project repositories
`project delete BRANCH`	deletes local and remote BRANCH in all project repositories
`project commit MESSAGE`	add changed/untracked files, commit them in project repos
`project push`	push all commits in all project repos, creates remote branches
`project merge [BRANCH]`	merges current branch of project repos into BRANCH
`project promote ENVIRONMENT`	promotes an environment branch to the next environment branch
`cli`	CLI script for common tasks (run within the container!)
`services/config/`	Application Configuration files for the services
`code/`	Symlinks to src repositories for the application development
`volumes/shared/src/`	Repos of all Tryton and collecting_society modules
`volumes/shared/ref/`	Repos of some pinned packages we use, just for reference

Data¶

`db-rebuild`	db-rebuild script for the database and demodata
`volumes/postgresql-data/`	Files of the postgres database
`volumes/echoprint-data/`	Files of the echoprint database
`volumes/shared/data/datasets/`	Demodata generation scripts for each tryton model
`volumes/shared/data/fingerprints/`	Ingestable demo fingerprints for echoprint
`volumes/shared/data/updloads/generate.sh`	Audiofile generation and compression script

Documentation¶

`docs-build`	docs-build script to build the project documentation
`docs/index.html`	Main index file of the built documentation

Tests¶

`service-test`	service-test script to run all service application tests
`tests/cover_*/index.html`	Html summary of coverage for webapi/webgui and worker
`tests/screenshots/`	Screenshots of the integration tests

Installation¶

To install the docker development environment from scratch, carry out the instructions of the following sections consecutively.

Requirements¶

Linux or OS X system
git
python >=3.7
pyyaml
docker >= 17.12.0
docker-compose >= 1.28.6

Repositories¶

In the first step, the repositories of the services have to be cloned and some filesystem preparation tasks have to be performed. Clone the collecting_society_docker repository into your working space:

$ cd MY/WORKING/SPACE
$ git clone https://github.com/C3S/collecting_society_docker.git

Switch to the root directory of the repository:

$ cd collecting_society_docker

Note

All setup and maintainance tasks are performed in the root path of this repository.

Checkout the Environments branch to build: development, staging, production:

$ git checkout <ENVIRONMENT>

If you just want to try out the software, the default development branch is recommended.

Copy the main environment variable example file .env.example to .env:

$ cp .env.example .env

Adjust the following variables:

Variable	Values	Default	Description
`DEBUGGER_DEBUGPY`	0\|1	0	Install debugpy during build process for debugging
`GIT_SSH`	0\|1	0	Checkout git repositories via ssh
`GIT_USER_NAME`	string	“”	Username for git commits (optional)
`GIT_USER_EMAIL`	string	“”	Email for git commits (optional)
`GIT_USER_SIGNINGKEY`	string	“”	16-hex-digit GPG key id for signed commits

Run the project script update command, which checkouts the service repositories, creates the service folders and copies the configuration example files (~5-10 minutes):

$ ./project update

Configuration¶

For staging and production environments:

Adjust the variables in .env (hostnames, ports, usernames, paths, etc).

Adjust the secrets:

File	Variable
`sevices/database.env`	`POSTGRES_PASSWORD`
`sevices/erpserver.env`	`PGPASSWORD` `TRYTON_EMAIL`
`sevices/webapi.env`	`PYRAMID_AUTHENTICATION_SECRET` `PYRAMID_SESSION_SECRET`
`sevices/webgui.env`	`PYRAMID_AUTHENTICATION_SECRET` `PYRAMID_SESSION_SECRET`
`sevices/worker.env`	`ECHOPRINT_TOKEN` `WORKER_PROTEUS_PASSWORD`
`services/config/collecting_society.<ENVIRONMENT>.conf`	`privatekey` `certificate` `super_pwd`
`services/config/collecting_society.passfile`	plaintext

Add basic http authentication, if needed:

$ sudo htpasswd -c volumes/nginx-htpasswd/collecting_society.test <USERNAME>
$ sudo ln -s collecting_society.test volumes/nginx-htpasswd/api.collecting_society.test

Images¶

Each service runs on a separate docker container. A docker container is a running instance of a prebuilt docker image. The docker images for all services need to be built first.

The initial build of the containers will take some time (around 15-30 minutes):

$ docker compose build

Database¶

After building the images, the services can be started. On the first run, the database and demodata is created (takes about 10 to 15 minutes):

$ docker compose up

The services should now be running and ready for clients to connect.

Domains¶

The webserver uses domain based routing of requests. In order to resolve the testing domains to localhost, add the following lines to /etc/hosts:

0.0.1   collecting_society.test
0.0.1   api.collecting_society.test
0.0.1   erp.collecting_society.test

Test the connection by following the instructions in Webbrowser Usage.

Email¶

For staging staging and production environments:

Find out the docker host ip of the docker0 interface (usually 172.17.0.1) and the docker subnet (usually 172.0.0.0/8):

$ ip -4 addr show docker0 | grep -Po 'inet \K[\d.]+'
172.17.0.1

$ docker network inspect collecting_society_frontend | grep -Po '(?<=Subnet": ").*(?=")'
172.25.0.0/16

Adjust the following .env variables:

# docker host ip
MAIL_HOST=172.17.0.1

# MTA port on host
MAIL_PORT=25

# activate mails
MAIL_TO_REAL_WORLD=1

Configure your MTA to listen to the docker host ip and to accept mails from the docker subnet. If you use exim4 for example, adjust /etc/exim4/update-exim4.conf.conf and restart the MTA:
```
# add docker host ip
dc_local_interfaces='127.0.0.1 ; ::1 ; 172.17.0.1'

# add docker subnet
dc_relay_nets='172.0.0.0/8'
```

Add an iptable rule to allow packages from the docker subnet to reach the host MTA service:

# iptables -A INPUT -s <SUBNET> -d <HOSTIP> -p tcp --dport 25 -j ACCEPT
iptables -A INPUT -s 172.0.0.0/8 -d 172.17.0.1 -p tcp --dport 25 -j ACCEPT

Tryton¶

To connect to Trytond, you can use one of the several Tryton client applications or APIs:

The Tryton Web Client is installed by default and ready for use.
For back-office use the Gtk based Tryton Desktop Client is recommended.

To install the desktop client, create a venv and install it using pip:

$ cd MY/WORKING/SPACE
$ python -m venv tryton
$ source tryton/bin/activate
$ pip install tryton~=7.0.0
$ ./tryton/bin/tryton

Note

The Trytond server and the Tryton client are required to have the same version branch.

Depending on the OS, there might be different ways to install the dependencies (see doc/installation.rst and tryton-client package of Ubuntu 21). Summary for Debian/Ubuntu:

apt-get install \
    gir1.2-gtk-3.0 \
    librsvg2-common \
    python3 \
    python3-cairo \
    python3-chardet \
    python3-dateutil \
    python3-gi \
    python3-gi-cairo \
    python3-pkg-resources

Test, if Tryton is running:

$ ./bin/tryton

Note

To get rid of the update check, click on the arrow down symbol in the upper right corner and uncheck “Check Version”.

For easy startup create a startup script:

Create the file /usr/local/bin/tryton in your prefered editor, e.g.:
```
$ sudo vim /usr/local/bin/tryton
```
Paste the following lines, set TRYTONPATH to the path of the tryton repository:
```
#!/bin/bash
./MY/WORKING/SPACE/tryton/bin/tryton
```
Set the execution flag of the script:
```
$ sudo chmod u+x /usr/local/bin/tryton
```
Test the script:
```
$ tryton
```

Test the connection by following the instructions in Tryton Usage.

Configuration¶

The services are configured via:

Project configuration: project.yml
Application environment: development, staging, production, testing
Global and service specific envvar files for the containers: .env, service/<SERVICE>.env
Application specific configuration files: *.conf, *.ini

Note

Sane defaults for a development setup are given and should work as provided, so this section might be skipped to start with development.

Note

Some files are tracked in git as FILE.example and are initally copied to the untracked FILE but not overwritten by the project script update command. The script will print notifications and diffs, when those files need to be changed manually.

Project¶

The project configuration file project.yml describes the tasks to perform to setup and update the environments. Tasks may include the the creation or copying of files, folders and symlinks and the checkout specific branches or tags of upstream and project repositories.

Environments¶

The services are configured differently for certain application environments. The differences on each level include:

docker: mapped ports, volume handling
database: demodata generation
application: debug switches, template caching

Context	Ports	Volumes	Demodata	Debug	Cache
`development`	all	local mounts	yes	on	off
`staging`	public	local mounts	yes	off	on
`production`	public	docker managed	no	off	on
`testing`	public	docker managed	no	off	on

For each of the environments except testing (all environments can be tested), there is a corresponding branch with the same name in this repository and most of the main subrepositories pre-configured for this environment.

Envvars¶

The .env file in the root path of the repository is the main envvar file and prefered place to specify configuration variables for all services. It is included in all main service containers. The variables might be overridden in a service container by the corresponding services/<SERVICE>.env.

The .env file is also processed by docker compose by convention and contains variables for the build process as well as for the project script.

See also

Compose CLI environment variables

.env¶

Variable	Values	Description
`PROJECT`	string	project name
`BRANCH`	string	branch of project repositories
`ENVIRONMENT`	“development” “staging” “production”	environment, switch for config files
`BUILD`	string	build number added by ci
`CHECKOUT_REFERENCES`	0\|1	checkout pip packages for reference
`COMPOSE_DOCKER_CLI_BUILD`	0\|1	use BuildKit for docker builds
`COMPOSE_PROJECT_NAME`	string	prefix for containers
`COMPOSE_IGNORE_ORPHANS`	0\|1	suppress orphan container warnings
`DEBUGGER_WINPDB`	0\|1	install packages for winpdb in images
`DEBUGGER_DEBUGPY`	0\|1	install packages for debugpy in images
`DEBUGGER_DEMODATA_WAIT`	0\|1	wait for debugger in demodata creation
`WORKDIR`	PATH	workdir for images
`GIT_SSH`	0\|1	use git via ssh
`GIT_USER_NAME`	string	set git username in repositories
`GIT_USER_EMAIL`	string	set git email in repositories
`GIT_USER_SIGNINGKEY`	string	GPG key for signing commits
`POSTGRES_HOSTNAME`	string	hostname of postgres server
`POSTGRES_PORT`	integer	port of postgres server
`TRYTON_HOSTNAME`	string	hostname of tryton server
`TRYTON_DATABASE`	string	name of the tryton database
`TRYTON_PORT`	integer	port of tryton server
`TRYTON_VERSION`	string	version of tryton to use
`VIRTUAL_HOST_WEBGUI`	URI	nginx URI for the webgui service
`VIRTUAL_PORT_WEBGUI`	integer	nginx reverse port for webgui
`VIRTUAL_HOST_WEBAPI`	URI	nginx URI for the webapi service
`VIRTUAL_PORT_WEBAPI`	integer	nginx reverse port for webapi
`VIRTUAL_HOST_ERPSERVER`	URI	nginx URI for the erpserver service
`VIRTUAL_PORT_ERPSERVER`	integer	nginx reverse port for erpserver
`MAIL_HOST`	string	hostname of the mail server
`MAIL_PORT`	integer	port of the mail server
`MAIL_DEFAULT_SENDER`	EMAIL	default sender email address
`MAIL_TO_REAL_WORLD`	0\|1	simulate sending mails or not
`PYRAMID_SCHEMA`	SCHEMA	schema of pyramid server
`PYRAMID_TRUSTED_PROXY`	IP	trusted IP for pyramid server
`WEBAPI_URL`	URL	URL of web api
`WEBAPI_CORS`	URL	allowed origins for web api CORS
`WEBAPI_VERSION`	string	version of web api
`WEBAPI_ENDPOINT_DATATABLES`	string	REST endpoint name for datatables
`WEBAPI_ENDPOINT_REPERTOIRE`	string	REST endpoint name for repertoire
`WEBAPI_CONTENT`	PATH	path to content folder (upload)
`WEBAPI_STORAGE`	PATH	path to storage folder (processing)
`ECHOPRINT_SCHEMA`	SCHEMA	schema of echoprint server
`ECHOPRINT_HOSTNAME`	string	hostname of echoprint server
`ECHOPRINT_PORT`	integer	port of echoprint server
`WORKER_PROTEUS_USER`	string	tryton username for proteus client
`WORKER_DISEMBODY_DROPPED_FILES`	“yes”\|”no”	delete upload content to save space

database¶

POSTGRES_PASSWORD

string

superuser password for postgresql

erpserver¶

`PGPASSWORD`	string	password for postgresql connections
`TRYTON_EMAIL`	string	email address of tryton admin user

webapi¶

`PYRAMID_AUTHENTICATION_SECRET`	string	secret for authentication
`PYRAMID_SESSION_SECRET`	string	secret for sessions

webgui¶

`PYRAMID_AUTHENTICATION_SECRET`	string	secret for authentication
`PYRAMID_SESSION_SECRET`	string	secret for sessions

worker¶

`ECHOPRINT_TOKEN`	string	authtoken for echoprint server
`WORKER_PROTEUS_PASSWORD`	string	tryton password for proteus client

Applications¶

The applications (trytond, proteus, pyramid) provide distinct files for all application environments, which are included depending on the value of the .env variable ENVIRONMENT. The applications might use envvars as well indicated by the syntax ${VARIABLE} in the configuration file. The same syntax can also be used in project.yml. The following sections provide a list of all envvar and configuration files for each application.

Trytond¶

Services: erpserver, webapi, webgui

`.env`	main envvar file
`services/config/collecting_society.<ENVIRONMENT>.conf`	trytond config
`services/config/collecting_society.passfile`	initial trytond admin password

Proteus¶

Services: worker

`.env`	main envvar file
`services/worker.env`	service envvar file
`services/config/collecting_society_worker.config.ini`	worker/proteus config

Pyramid¶

Services: webapi, webgui

`.env`	main envvar file
`services/web[api\|gui].env`	service envvar file
`services/config/portal_web.<ENVIRONMENT>.ini`	pyramid config
`services/config/collecting_society_web.<ENVIRONMENT>.ini`	pyramid plugin config

Usage¶

There are several ways to interact with the services:

The docker compose CLI is the prefered general high level docker tool for everyday use.
The docker CLI provides sometimes more useful low level commands.
The Scripts in the root folder are provided for comfort or automatisation.
The CLI script provides special maintainance commands for the services (for use within the containers).

If you tend to forget the commands or syntax, try getting used to the help commands:

List docker compose commands	`docker compose --help`
Help for docker compose command	`docker compose COMMAND --help`
List docker commands	`docker --help`
Help for docker command	`docker COMMAND --help`
List scripts	`ls -F \| grep '*$'`
Help for scripts	`./SCRIPT --help`
List CLI command	`docker compose [exec\|run --rm] erpserver --help`
Help for CLI command	`docker compose [exec\|run --rm] erpserver COMMAND --help`

Run¶

Start services	`docker compose up`
Start services in the background	`docker compose up -d`
Start a certain service (in the background)	`docker compose up SERVICE [-d]`
Run a command on a running\|new container	`docker compose [exec\|run --rm] SERVICE CMD`
Run CLI command on a running\|new container	`docker compose [exec\|run --rm] SERVICE [cli] CMD`
Open a shell on a running\|new container	`docker compose [exec\|run --rm] SERVICE bash`
Run CLI command inside a container shell	`[cli] CMD`
Build documentation	`./docs-build`
Run tests	`./service-test`
Scale services on demand	`docker compose scale SERVICE=#`
Stop services	`docker compose stop`
Stop a certain service	`docker compose stop SERVICE`
Stop and remove containers/volumes/networks	`docker compose down`

See also

[SERVICE]: Table of Services, [CMD]: CLI.

Note

Always prefer exec to run --rm, if containers are already running.

Update¶

Update repositories	`./project update`
Diff repos/example files	`./project diff`
Build images	`docker compose build`
Update database	`docker compose [exec\|run --rm] erpserver db-update`

Update the repositories/files/folders:
```
$ ./project update
```
The script will print notifications and instruction, if further steps are neccessary.

Note

The project script update command will also try to update the collecting_society_docker repository and thus itself first, before updating the subordinate repositories.
If there were changes to the *.example files, diff the files and apply changes manually:
```
$ ./project diff
```
If there were changes in the Dockerfile, rebuild all docker images:
```
$ docker compose build
```
If you run into problems, you can also rebuild all docker images without cache. Just remove all project images (also the dangling ones) before the execution of the build command.
If there were changes in the collection_society repository, update the database:
```
$ docker compose run --rm erpserver db-update
```
If you run into problems and don’t care about the data, you can also recreate the database:
```
$ ./db-rebuild
```

Inspect¶

Attach to the logs of a certain service	`docker compose logs [-f] SERVICE`
Open a shell on a service container	`docker compose run --rm SERVICE bash`
Open a shell on a running container	`docker compose exec bash`
List project docker containers	`docker compose ps`
List project docker images	`docker compose images`
List project docker containers	`docker compose ps [-a]`
List processes of project container	`docker compose top`
Show used resources for containers	`docker stats`
List docker images	`docker images ls [-a]`
List docker networks	`docker network ls`
List docker volumes	`docker volume ls`
Inspect a container/volume/network/…	`docker inspect ID\|NAME`

Remove¶

Warning

The docker commands apply to all docker containers on the host.

Remove project containers/networks/volumes	`docker compose down`
Remove all stopped docker containers	`docker container prune`
Remove all dangling images to free diskspace	`docker image prune`
Remove volumes	`docker volume rm VOLUMENAME`

Note

For VOLUMENAME see the output of docker volume ls.

Remove all containers, networks, volumes and images:

$ docker compose -f docker-compose.documentation.yml down -v --rmi all
$ docker compose -f docker-compose.testing.yml down -v --rmi all
$ docker compose down -v --rmi all
$ docker image prune

Note

The multiple down commands are needed, as testing and documentation have separate containers, but are based on the same multistage Dockerfile.

Database¶

Create	`docker compose [exec\|run --rm] erpserver db-create [NAME]`
Copy	`docker compose [exec\|run --rm] erpserver db-copy [--force] [SOURCENAME] [TARGETNAME]`
Backup	`docker compose [exec\|run --rm] erpserver db-backup [NAME] > /shared/tmp/db.backup`
Delete	`docker compose [exec\|run --rm] erpserver db-delete [NAME]`
Setup	`docker compose [exec\|run --rm] erpserver db-setup [NAME]`
Rebuild	`docker compose [exec\|run --rm] erpserver db-rebuild [NAME]` `./db-rebuild`
Examine	`docker compose run --rm erpserver db-connect [NAME]`
Console	`docker compose run --rm erpserver db-console [NAME]`

Note

[NAME] is optional and defaults to collecting_society.

Note

If the setup/rebuild hangs, look for and delete the ./volumes/shared/running_db_creation.delete_me locking file.

The database files are stored in ./volumes/postgresql-data. If the postgres setup itself seem to be broken, you can delete and recreate the folder:

$ docker compose down
$ sudo rm -rf ./volumes/postgresql-data/
$ docker compose up

Warning

All data in this database will be deleted!

Note

The uid/gid of the folder and files matches those of the postgres user in the cointainer, so sudo is probably neccessary to be able to delete them.

Scripts¶

The scripts are either intended to make some operations more comfortable or for automatisation using a build server (CI). The following sections contain a brief synopsis about each of the provided scripts as provided by the --help option. The usual syntax is object-operation.

project¶

$ ./project --help
usage: ./project

Performs development and maintainance tasks for the project.

optional arguments:
  -h, --help         show this help message and exit

subcommands:
  (default: status)
    update           Updates files, folders, symlinks and repos
    status           Prints the git status of the project repositories
    diff             Prints the diff of the project repos and the example files
    pull             Pulls the current branch for all project repositories
    checkout         Checksout a branch in all project repositories
    delete           Deletes a local and remote branch in all project repos
    commit           Commits changes and untracked files to the project repositories
    push             Pushes commits in all project repos, creates missing remote branches
    merge            Merges the current branch into another branch in all project repos
    promote          Merges an environment branch into the next stage environment branch

$ ./project update --help
usage: ./project update [-h] [-v] [--branch NAME] [--environment NAME] [--reset] [--ci]

Updates files, folders, symlinks and repos.

optional arguments:
  -h, --help          show this help message and exit
  -v, --verbose       verbose output, -vv for debug output
  --branch NAME       Branch name (default: .env [feature-updatescript])
  --environment NAME  Environment name (default: .env [development])
  --reset             overwrites the configuration files with example files (default: False)
  --ci                continues integration mode: reset, debug, colorless (default: False)

$ ./project status --help
usage: ./project status [-h] [-v]

Prints the git status of the project repositories.

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

$ ./project diff --help
usage: ./project diff [-h] [-v]

Prints the diff of the project repos and the example files.

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

$ ./project pull --help
usage: ./project pull [-h] [-v]

Pulls the current branch for all project repositories.

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

$ ./project checkout --help
usage: ./project checkout [-h] [-v] [BRANCH]

Checksout a branch in all project repositories.

positional arguments:
  BRANCH         Branch name (default: checkedout [feature-updatescript])

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

$ ./project delete --help
usage: ./project delete [-h] [-v] [-f] [--no-local-delete] [--no-remote-delete] BRANCH

Deletes a local and remote branch in all project repos.

positional arguments:
  BRANCH              Branch name

optional arguments:
  -h, --help          show this help message and exit
  -v, --verbose       verbose output, -vv for debug output
  -f, --force         Force deletion of not fully merged branches (default: False)
  --no-local-delete   Don't delete local branch (default: False)
  --no-remote-delete  Don't delete remote branch (default: False)

$ ./project commit --help
usage: ./project commit [-h] [-v] MESSAGE

Commits changes and untracked files to the project repositories.

positional arguments:
  MESSAGE        Commit message

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

$ ./project push --help
usage: ./project push [-h] [-v]

Pushes commits in all project repos, creates missing remote branches.

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

$ ./project merge --help
usage: ./project merge [-h] [-v] [-f] [--no-delete] [--no-local-delete] [--no-remote-delete]
                       [--no-push] [BRANCH]

Merges the current branch into another branch in all project repos.

positional arguments:
  BRANCH              Target branch name (default: development)

optional arguments:
  -h, --help          show this help message and exit
  -v, --verbose       verbose output, -vv for debug output
  -f, --force         Force deletion of not fully merged branches (default: False)
  --no-delete         Don't delete branch after merge (default: False)
  --no-local-delete   Don't delete local branch after merge (default: False)
  --no-remote-delete  Don't delete remote branch after merge (default: False)
  --no-push           Don't push branch after merge (default: False)

$ ./project promote --help
usage: ./project promote [-h] [-v] ENVIRONMENT

Merges an environment branch into the next stage environment branch.

positional arguments:
  ENVIRONMENT    Environment to be promoted to the next stage

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  verbose output, -vv for debug output

service-test¶

$ ./service-test --help
Usage: ./service-test [service] [--down] [--build] [--keep] [--lint]
                  [--ci] [--ci-branch NAME] [--ci-environment NAME]
                  [--help] [PARAMS]

  This script runs the unit/function/integration tests and linter for the services:
    - erpserver (tryton)
    - web (pyramid)
    - worker (echoprint)

Options:
  service: web|worker|erpserver|all (default: all)
  --down: immediately stop, remove the container and exit
  --build: build images and recreate the test database template
  --keep: keep container running
  --lint: only lint the code, don't run the tests
  --ci: continous integration mode
        - update repositories (overrides config files!)
        - build images
        - recreate the test database template
        - run tests and linter
        - stop and remove the container
  --ci-branch: branch to test
  --ci-environment: environment to test
  --help: display this help
  PARAMS: are passed to pytest

docs-build¶

$ ./docs-build --help
Usage: ./docs-build [--down] [--build] [--keep] [--no-autoapi]
                    [--ci] [--ci-branch NAME] [--ci-environment NAME]
                    [--help]

  This script builds the documentation with sphinx.

Options:
  --down: immediately stop and remove the container and exit
  --build: build images
  --keep: keep container running
  --no-autoapi: don't parse the modules
  --ci: continous integration mode
        - update repositories (overrides config files!)
        - build images
        - build docs
        - stop and remove the container
  --ci-branch: branch to test
  --ci-environment: environment to test
  --help: display this help

db-rebuild¶

$ ./db-rebuild --help
Usage: ./db-rebuild [--ci] [--help]

  This script deletes and recreates the database and generates the demodata.

Options:
  --ci: stops the services before, starts the services detached afterwards
  --help: display this help

CLI¶

The ./volumes/shared/cli script contains a CLI for special service maintainance commands. Within the containers it is available in the working directory /shared/cli. For convenience and to ensure the same command invokation syntax of exec and run --rm, the commands of the script are also available directy via /shared/COMMAND.

Warning

All CLI commands should only be executed within a service container!

Note

Not all commands will work on any service.

Usage:

On the host:

$ docker compose run --rm SERVICE COMMAND
$ docker compose exec SERVICE COMMAND

For example:

$ docker compose run --rm erpserver db-rebuild
$ docker compose exec erpserver db-rebuild

Note

Use exec if the container is already running, e.g. in another terminal window after a docker compose up. Use run --rm if no container is running and your just want to start it for a single task upon which it is removed again (-rm). To start more than a single task, you would want to ‘go inside a container’ by running a bash command, e.g. docker compose run --rm erpserver bash.

Inside a service container:

$ COMMAND

For example:

$ db-rebuild

Help:

$ cli --help
$ COMMAND --help

Commands:

$ cli --help
Usage: cli [OPTIONS] COMMAND [ARGS]...

  Command line interface to setup and maintain services in docker
  containers.

Options:
  --help  Show this message and exit.

Commands:
  db-backup            Dumps the postgres database DBNAME to stdout.
  db-connect           Opens a SQL console for the database DBNAME.
  db-console           Opens trytond console.
  db-copy              Creates the postrges database DBNAME_DST from...
  db-create            Creates the postrges database DBNAME.
  db-delete            Deletes the postrges database DBNAME.
  db-rebuild           Deletes DBNAME and executes db setup
  db-setup             Creates and sets up the postgres database...
  db-update            Updates tryton modules for database DBNAME.
  docs-build           Builds the Sphinx documentation.
  pip-install          Installs required packages for a SERVICE with...
  service-deploy       Deploys the services (erpserver, webgui,...
  service-healthcheck  Healthcheck for the services.
  service-lint         Runs linter for a service (erpserver,...
  service-test         Runs all tests for a service (erpserver, web,...

db-backup¶

$ db-backup --help
Usage: cli db-backup [OPTIONS] [DBNAME]

  Dumps the postgres database DBNAME to stdout.

Options:
  --help  Show this message and exit.

db-connect¶

$ db-connect --help
Usage: cli db-connect [OPTIONS] [DBNAME]

  Opens a SQL console for the database DBNAME.

Options:
  --help  Show this message and exit.

db-console¶

$ db-console --help
Usage: cli [OPTIONS] COMMAND [ARGS]...

  Command line interface to setup and maintain services in docker containers.

Options:
  --help  Show this message and exit.

db-copy¶

$ db-copy --help
Usage: cli db-copy [OPTIONS] DBNAME_SRC DBNAME_DST

  Creates the postrges database DBNAME_DST from template DBNAME_SRC.

Options:
  --force / --no-force  Force execution (default: no)
  --help                Show this message and exit.

db-create¶

$ db-create --help
Usage: cli db-create [OPTIONS] [DBNAME]

  Creates the postrges database DBNAME.

  The execution is skipped if the database already exists.

Options:
  --help  Show this message and exit.

db-delete¶

$ db-delete --help
Usage: cli db-delete [OPTIONS] [DBNAME]

  Deletes the postrges database DBNAME.

  On error the deletion is retried several times.

Options:
  --help  Show this message and exit.

db-rebuild¶

$ db-rebuild --help
Usage: cli db-rebuild [OPTIONS] [DBNAME]

  Deletes DBNAME and executes db setup

Options:
  -r, --reclimit INTEGER      Maximum numbers of objects (default: 0 = all)
  -d, --dataset TEXT          dataset in ./data/datasets/ to generate
                              (default: all)
                              can be used multiple times
  -e, --exclude TEXT          datasets in ./data/datasets/ to exclude
                              (default: none)
                              can be used multiple times
  --template / --no-template  Use template db for dataset deps (default: yes)
  --cache / --no-cache        Use/Recreate template db for dataset deps
                              (default: no)
  --pdb / --no-pdb            Start pdb on error (default: no)
  --help                      Show this message and exit.

db-setup¶

$ db-setup --help
Usage: cli db-setup [OPTIONS] [DBNAME]

  Creates and sets up the postgres database DBNAME.

  The execution is skipped if the database already exists. The execution
  might be forced (omits the db creation, if it exists).

  Generates production and demodata.

  During installation a lockfile is created on the host to prevent multiple
  execution from different docker containers.

Options:
  -r, --reclimit INTEGER      Maximum numbers of objects (default: 0 = all)
  -d, --dataset TEXT          dataset in ./data/datasets/ to generate
                              (default: all)
                              can be used multiple times
  -e, --exclude TEXT          datasets in ./data/datasets/ to exclude
                              (default: none)
                              can be used multiple times
  --template / --no-template  Use template db for dataset deps (default: yes)
  --cache / --no-cache        Regenerate template db for dataset deps
                              (default: no)
  --force / --no-force        Force execution (default: no)
  --pdb / --no-pdb            Start pdb on error (default: no)
  --help                      Show this message and exit.

db-update¶

$ db-update --help
Usage: cli db-update [OPTIONS] [TRYTONDCONF] [DBNAME]

  Updates tryton modules for database DBNAME.

  Modules can be provided, default is 'collecting_society'. If modules are
  'all', all modules are updated.

Options:
  -m, --modules TEXT  Single module or comma separated list of modules to
                      update. Whitspace not allowed!
  --help              Show this message and exit.

docs-build¶

$ docs-build --help
Usage: cli docs-build [OPTIONS]

  Builds the Sphinx documentation.

  Installs pip packages of all modules so they can be found by Sphinx.
  autoapi and Sphinx are started with docs/build.sh.

Options:
  --autoapi / --no-autoapi  Activate autoapi (default: yes)
  --help                    Show this message and exit.

pip-install¶

$ pip-install --help
Usage: cli pip-install [OPTIONS] [SERVICE]

  Installs required packages for a SERVICE with pip.

  Requirements have to be defined in `./shared/config/pip/SERVICE.pip`.

  After installation a flag file is created within the container to avoid
  multiple execution during its lifespan.

Options:
  --help  Show this message and exit.

service-deploy¶

$ service-deploy --help
Usage: cli service-deploy [OPTIONS] [SERVICE]

  Deploys the services (erpserver, webgui, webapi, worker, fingerprint).

  Installs pip packages, creates and sets up database and runs the
  application.

Options:
  --help  Show this message and exit.

service-healthcheck¶

$ service-healthcheck --help
Usage: cli service-healthcheck [OPTIONS] [SERVICE]

  Healthcheck for the services.

Options:
  --help  Show this message and exit.

service-lint¶

$ service-lint --help
Usage: cli service-lint [OPTIONS] [SERVICE]

  Runs linter for a service (erpserver, web/webgui/webapi, worker).

  If PATH is provided, only the path is linted, not the service. If SERVICE
  is 'all', all services are linted.

Options:
  --path TEXT  Custom path with files to lint
  --help       Show this message and exit.

service-test¶

$ service-test --help
Usage: cli service-test [OPTIONS] [SERVICE] [NARGS]...

  Runs all tests for a service (erpserver, web, worker).

  Starts pytests and prints output to stdout.

  Creates the test database template DBNAME_template, if not existant. On
  RESET, the database DBNAME will be recreated from this template and the
  temporary tryton file folder will be deleted.

  The location of the temporary tryton upload folder is configured in
  `./shared/config/trytond/testing_DBTYPE.conf` (currently
  `./shared/tmp/files`).

  The location of the screenshots of integration tests is configured within
  `<portal_web>/tests/config.py` (currenty `./shared/tmp/screenshots).

  The PATH to tests may be defined to test certain testfiles, testclasses or
  test methods (see pytest for the syntax). If no PATH is given, all tests
  of portal_web and plugins are included. The test files should be stored
  below the following subpaths by convention:

      <portal_web||plugin>/tests/unit (unittest)

      <portal_web||plugin>/tests/functional (webtest)

      <portal_web||plugin>/tests/integration (selenium)

  Additional NARGS will be passed to pytests.

Options:
  --dbname TEXT         Name of database (default: test)
  --reset / --no-reset  Reset the database (default: yes)
  --path TEXT           Searchpath for tests (see pytest)
  --help                Show this message and exit.

Tryton¶

Web Client¶

Open the webbrowser and point it to the

tryton web client: http://erp.collecting_society.test

Choose the database collecting_society and login as user:

Username	Password	Roles
`admin`	`password`	Admin
`storehouse001`	`password`	Storehouse Admin

Desktop Client¶

Start Tryton:

$ tryton

Note

The Tryton client configuration files are stored in ~/.config/tryton/7.0/.

Open a connection to Trytond:

host	`erp.collecting_society.test:80`
database	`collecting_society`

Login as user:

Username	Password	Roles
`admin`	`password`	Admin
`storehouse001`	`password`	Storehouse Admin

See also

Tryton Usage Documentation

The database entries can be found in the navigation tree:

Collecting Society: Societies, Tariffs, Allocations, Distributions
Licenser: Artists, Releases, Creations, Licenses, Labels, Publishers
Licensee: Events, Locations, Websites, Releases, Devices, Declarations, Utilisations
Portal: Access
Archiving: Storehouses, Harddisks, Filesystems, Contents

Other important entries are:

Party: Parties, Addresses
Administration / Users: Users, Web Users
Administration / Sequences: Sequences

Webbrowser¶

Open the webbrowser and point it to the

webgui: http://collecting_society.test
webapi: http://api.collecting_society.test

Login as demo user:

Username	Password	Roles
`allroles1@collecting-society.test`	`password`	Licenser, Licensee
`licenser1@collecting-society.test`	`password`	Licenser
`licensee1@collecting-society.test`	`password`	Licensee

Development¶

Environment¶

Project¶

The tasks to setup each environment can be configured in ./project.yml:

<ENVIRONMENT>:

  commands:
    <COMMAND>: {}

  tasks:
    <COMMAND>:

      - name: <NAME>
        actions: [<ACTION>, <ACTION>]
        <KEY>: <VALUE>

      - name: <NAME>
        actions: [<ACTION>, <ACTION>]
        <KEY>: <VALUE>
        batch:
          - name: <NAME>
            <KEY>: <VALUE>

  actions:
    <ACTION>: {}
    <ACTION>: []

Key	Description
`ENVIRONMENT`	environment, for which the tasks are performed. inheritance: production -> staging -> testing -> development
`commands`	configuration variables availbable for each command
`COMMAND`	main commands maps to `@command` functions in the project script
`tasks`	list of tasks to perform consecutively for each command
`NAME`	name of the task, required for all tasks
`ACTION`	actions to perform consecutively for each task, maps to `@action` functions in the project script
`actions`	[dictionary] configuration values available in actions [list] action group with actions to perform consecutively

Commands can be invoked via the project script. For available commands, see the @command decorated functions in the script.

Each command processes its task list and for each task the defined actions consecutively. Each action receives the task dictionary and expects the task to have the proper key/value pairs (e.g. repos need a source, etc). The command/action config dictionary is also available to the actions and might configure how the action should be performed. For available actions, see the @action decorated functions in the script.

For a list of command/action configuration variables, see the comments in ./project.yml.

Batch tasks will use the key/value pairs of its parent updated with itself. In inherited environments, tasks may be changed by using the same name of the inherited task.

Branches¶

Each project repository has a branch for all environments. To switch a branch for all project repositories:

$ ./project checkout BRANCH

Using feature branches is encouraged. To create a new local feature branch for all repositories:

development$ ./project checkout feature-<FEATURENAME>

The basic workflow:

Create a feature branch. Remote branches are always prefered during checkout:
```
development$ ./project checkout feature-branch
```
Develop the code:
```
feature-branch$ [...]
```
Test the code:
```
feature-branch$ ./service-test
```

Check the status of the workdirs:

feature-branch$ ./project status
feature-branch$ ./project diff [-v]

Commit the changes and new files:

feature-branch$ ./project commit "commit message"

Push the branch, if the feature branch should be shared:
```
feature-branch$ ./project push
```
Delete the branch, if the feature branch should be discarded. Both the local and remote branch will be deleted:
```
feature-branch$ ./project chechkout development
development$ ./project delete feature-branch
```
Merge the branch into development, when the feature is finished. This will delete the local and remote branch after the merge:
```
feature-branch$ ./project merge
```

Docker¶

Compose¶

The project consists of 3 separate docker compose setups:

Development/Staging/Production

Purpose: Main development/production setup of the services
Files
- docker compose.yml: main file
- docker compose.override.yml: override file, symlink to environment config (ports, volumes)
  - docker compose.development.yml: additions for development environment
  - docker compose.staging.yml: additions for staging environment
  - docker compose.production.yml: additions for productions environment
Usage: docker compose COMMAND
Services: Table of Services

Note

The docker-compose.override.yml is a docker compose convention.

Testing

Purpose: Manual/Automated testing, CI
Files
- docker-compose.testing.yml
Usage: docker compose -f docker-compose.testing.yml COMMAND
Services
- test_database: same as database
- test_erpserver: same as erpserver
- test_web: webapi + webgui
- test_worker: same as worker
- test_fingerprint: same as fingerprint
- test_browser: selenium

Documentation

Purpose: Manual/Automated builds of the documentation
Files
- docker-compose.documentation.yml
Usage: docker compose -f docker-compose.documentation.yml COMMAND
Services:
- documentation: sphinx build container

For more information, look into the docker-compose*.yml files.

Images¶

All images for all 3 docker compose setups are based on the same Dockerfile, which is located in ./services/build/Dockerfile. The key concepts for this image setup are:

Some and only those images not intended for production use are imported from Dockerhub (nginx, postgres, selenium).
All custom built images are based on Debian.
It is a multistage build. This means, that all intermediate stages can be reused for multiple images, leading to a stage hierarchy tree.
There are 2 branches in the tree:
- The compile branch contains the libraries needed for the compilation of the packages/applications.
- The service branch contains only the runtime dependencies for the packages/applications.
The packages/applications are compiled on images of the compile branch and in the end copied to the images on the service branch, which are used for the actual services.
Each image stage has 4 substages for the different environments:
- The production substage contains only the minimum of packages needed.
- The staging substage adds packages for staging.
- The testing substage adds packages for tests/CI/documentation.
- The development substage adds packages to develop comfortably.
The reason for both the division of compile/service branches as well as the substages matching the environment is to have slimmer images, smaller attack surfaces and a faster build time.
All images based on python use volumes/shared/docker-entrypoint.sh as entrypoint to detect and execute CLI commands provided by the volumes/shared/cli script.

The tree of the stages of the service branch (without substages):

                          base
                           |
                         python
          _________________|__________________
         |           |           |            |
      trytond     proteus    echoprint     compile
       |   |         |           |            |
erpserver webapi   worker   fingerprint  documentation
           |
          webgui

The tree of the stages of the compile branch (without substages):

                        base                       node
                         |                          |
                       python                   sao_compiled
                         |
                      compile
                         |
                   python_compiled
        _________________|____________________
       |                 |                    |
trytond_compiled   proteus_compiled   echoprint_compiled
       |
pyramid_compiled

The copy relations:

Image	Copy Sources
erpserver	trytond_compiled sao_compiled
webapi	pyramid_compiled
webgui	pyramid_compiled
worker	proteus_compiled
fingerprint	echoprint_compiled
documentation	trytond_compiled pyramid_compiled worker_compiled

Packages¶

This setup maintains three levels of package inclusion:

Debian packages

Python packages installed with pip

Source repositories for development purposes

Debian¶

The Debian packages installed for the applications can be found in the Dockerfile and are pinned, where reasonable. For a list of packages, search for apt-get install in ./services/build/Dockerfile.

Pip¶

The pip packages installed for the applications also can be found in the Dockerfile and are all pinned. For a list of packages, search for pip install in ./services/build/Dockerfile.

If CHECKOUT_REFERENCES is set to 1 in .env, the source code of those packages can also be found in the folder ./volumes/shared/ref/ and are provided for reference and for quick lookups during development. The source code is not used though. The repositories are cloned on the project script update command and can be configured in ./project.yml:

development:
  tasks:
    update:
      - name: checkout repos of pinned pip packages for reference
        batch:
          - name: <REPOFOLDER>
            source: <REPOSOURCE>
            version: tags/<TAG>

Repositories¶

Those packages, which are either under development or need to be updated regulary are git cloned into the folder ./volumes/shared/src/. Those packages are pip installed during runtime each time a container is started. The list of package requirements for each service container can be found in ./services/pip/<SERVICE>.pip.

The repositories are cloned and updated on each run of the project script update/pull command and can be configured in ./project.yml:

production:
  tasks:
    update:
      - name: update project repos
        batch:
          - name: <REPOFOLDER>
            source: <REPOSOURCE>
      - name: update upstream repos
        batch:
          - name: <REPOFOLDER>
            source: <REPOSOURCE>
            version: <BRANCH>

Services¶

To start all services with stdin attached to the service logs, use:

$ docker compose up

To start all services detached:

$ docker compose up -d

If you want to start only a certain service with its dependencies, use:

$ docker compose run --rm --service-ports SERVICE    service-deploy
  '---------------------------------------------'    '-------------'
                  host command                      container command

$ docker compose run --rm --service-ports webgui     service-deploy
$ docker compose run --rm --service-ports webapi     service-deploy
$ docker compose run --rm --service-ports erpserver  service-deploy

The host command explained:

docker compose run: Run a one-off command in a new container

--rm: The run command won’t remove the stopped container by default, so that it can be inspected after the run. To prevent the aggregation of stopped container states, this switch is recommended.

--service-ports: The run command is intended to be used, while the services are already running and does not map the service ports by default to prevent the port being allocated twice. This switch is used to enable the mapping of the service ports.

SERVICE: The service on which the command is executed

The container command explained:

service-deploy: The service-deploy CLI command to start the application

Note

The deploy scripts can be found in services/deploy/SERVICE.

To open a shell on a new container:

$ docker compose run --rm [--service-ports] SERVICE bash

Warning

Manual changes are not persisted when the container is stopped.

To open a shell on a running container:

$ docker compose exec SERVICE bash

Trytond¶

For the development of tryton modules it is recommended to open two shells within the erpserver:

One shell is to start the trytond server manually, as it often needs to be restarted.
The other shell is for the database update command to apply the changes to the database.

Start the first terminal, open a bash in the erpserver and start trytond:

$ docker compose run --rm --service-ports erpserver bash
> service-deploy

To restart the trytond server:

> <Ctrl+c>
> service-deploy

Start the second terminal, open another bash in the running container:
```
$ docker compose exec erpserver bash
```
To update the collecting_society module for the database:
```
> db-update
```
To update all modules for the database:
```
> db-update -m all
```

To start a trytond console (interactive python console with pool initialized):

> db-console

To connect to Trytond with the Tryton client, see Tryton Usage.

Note

Start Tryton with the -d/--debug flag to disable caching and -v and -l DEBUG for more verbose output, but expect a heavy slow down with the -d flag.

You can now start coding:

`code/collecting_society/`	trytond main module
`services/config/collecting_society.*`	trytond server config files
`~/.config/tryton/7.0/`	tyton client config files
`volumes/shared/src/`	all trytond module repositories
`volumes/trytond-files/`	trytond file storage

Lint the code:

docker compose exec erpserver service-lint

Pyramid¶

For the development of the pyramid application, it is sufficiant to just start all services with stdin attached to the service logs:

$ docker compose up

The application will monitor changes to files and restart itself automatically. You can now start coding:

`code/portal_web/`	pyramid main application code
`code/collecting_society_web/`	pyramid plugin code
`services/config/portal_web.*`	pyramid main application config files
`services/config/collecting_society_web.*`	pyramid plugin config files
`volumes/shared/ref/`	pinned python package repos for reference
`volumes/shared/tmp/logs`	log folder for some debugging flags
`volumes/shared/tmp/session`	cookie session data files
`volumes/shared/tmp/upload`	upload folder for audio/pdfs

Lint the code:

docker compose exec webgui service-lint

Debugging¶

Pdb¶

Pdbpp ist installed in all images with python installed and should work out of the box. Just add the line in the python file:

import pdb; pdb.set_trace()

If you want to debug a service, you need to start the service via the run command to attach stdin/stdout and add the --service-port flag:

$ docker compose run --rm --service-ports SERVICE service-deploy

If you want to debug application tests, you can add the --pdb flag to the service-test script or the service-test CLI command to jump into pdb on errors automatically.

If you want to debug the demodata generation, you can add the --pdb flag to the db-rebuild CLI command to jump into pdb on errors automatically.

Debugpy¶

If you use Visual Studio Code as your editor, you would want to install the Remote Containers extension, so you can work directly in the docker containers, including source level debugging from within VS Code. Just make sure that the environment variables in .env have the right values:

ENVIRONMENT=development
DEBUGGER_DEBUGPY=1

Now rebuild the docker images for the packages to be installed, cd to collecting_society_docker and start VSCode with "code .". The necessary .devcontainer.json and launch.json files are already included in the repositories.

To start debugging a container, click on the toast notification that will come up in the bottom right corner or click on the green field in the lower left corner of VS Code and select Remote-Containers: Reopen in Container. Then make sure the Python extension is installed in the container’s VS Code instance and reload, if necessary. Git History and GitLens are recommended but will require you to "apt-get install git" in the container. To start debugging, press Ctrl-Shift-D to open the debug sidebar and select the debug configuration in the drop-down box on the top, e.g. ‘Portal Attach’ (Settings for attaching the container can be adjusted in the file ./volumes/shared/.vscode/launch.settings). Press the play button left to the debug config drop-down box and a debug toolbar should appear.

Note

If you wish to debug other containers besides the default webgui, e.g. webapi or worker, change the service entry in .devcontainer.json accordingly, otherwise you will experience ‘connection refused’ errors. The service entry in .devcontainer.json will determine which container is being selected by the Remote-Containers plugin.

To debug demodata creation, you would want the debugger to wait until you attach. Use this env var:

DEBUGGER_DEMODATA_WAIT=1

Winpdb¶

To allow the winpdb debugger to attach to a portal script, make sure that the environment variables in .env have the right values:

ENVIRONMENT=development
DEBUGGER_WINPDB=1

Now rebuild the docker images for the packages to be installed an in your python file insert:

import rpdb2; rpdb2.start_embedded_debugger("password", fAllowRemote = True)

Make sure to open a port for the remote debugger in docker-compose.development.yml:

ports:
  - "52000:52000"

Install winpdb also outside the container and run it:

$ sudo apt-get install -y winpdb
$ winpdb

The processing container can be setup for debugging the same way. Make sure to only enable either of the both containers for debugging, not both the same time.

Trytond Console¶

Tryton can start an interactive python console with the pool initialized:

$ docker compose run --rm erpserver db-console

Postgres¶

Connect to postgres via console:

docker compose run --rm --service-ports erpserver bash
> db-connect

Connect to postgres via client (e.g. dbeaver)

Host: localhost
Port: 5432
User: postgres
Password: s0secret!!
Database: collecting_society

Print activity:

select pid, client_addr, wait_event_type, state, query from pg_catalog.pg_stat_activity;

Log all statements

Adjust the postgresql config file ./volumes/postgresql-data/postgresql.conf:

logging_collector = on
log_destination = 'csvlog'
log_directory = 'pg_log'
log_filename = 'postgresql.log'
log_statement = 'all'

Restart the database service:
```
docker compose restart database
```

Watch logfile:

sudo tail -f ./volumes/postgresql-data/pg_log/postgresql.log

Tests¶

The tests are performed on separate containers. To build the images on the first run, use the --build flag of the service-test script:

$ ./service-test --build

Run tests for all services (web, erpserver, worker):

$ ./service-test

If you develop the tests and need to start them more than once, you can use the --keep flag, to keep the container running and use the command multiple times:

$ ./service-test --keep

To stop and remove the container, when you have finished, enter

$ ./service-test --down

Note

All commits pushed to all C3S GitHub repositories are automatically CI tested with jenkins (needs authentication) using the same test script.

Trytond¶

Run all trytond tests (module tests, scenario doctests) once:

$ ./service-test erpserver

Run all trytond tests and keep the container running for the next test run:

$ ./service-test erpserver --keep

Stop the container afterwards:

$ ./service-test --down

You can append the normal pytest parameters:

$ ./service-test erpserver --keep [--path PATH] [PARAMETER]

Run all tests quietly, drop into pdb on errors, don’t suppress output:
```
$ ./service-test erpserver --keep --quiet --pdb -s
```

Run a tests matching a substring:

$ ./service-test erpserver --keep -k SUBSTRING

If you prefer, you can also execute the commands above from within the container:

$ docker compose -f docker-compose.testing.yml up -d
$ docker compose -f docker-compose.testing.yml exec test_erpserver bash

    # setup container
    > pip-install

    # run tests
    > service-test

    # exit container
    > exit

$ docker compose -f docker-compose.testing.yml down

Worker¶

Run all worker tests (module tests, scenario doctests) once:

$ ./service-test worker

Run all trytond tests and keep the container running for the next test run:

$ ./service-test worker --keep

Stop the container afterwards:

$ ./service-test --down

Note

The following commands will use the --keep flag by default. It will highly speed up the execution time, if you run the tests more than once.

You can append the normal pytest parameters:

$ ./service-test worker --keep [--path PATH] [PARAMETER]

Run all tests quietly, drop into pdb on errors, don’t suppress output:
```
$ ./service-test worker --keep --quiet --pdb -s
```

Run a specific set of tests:

$ ./service-test worker --keep --path PATH[/FILE[:CLASS[.METHOD]]]

For example:

$ TESTPATH=code/collecting_society_worker/collecting_society_worker/tests

$ ./service-test worker --keep \
    --path $TESTPATH/integration
$ ./service-test worker --keep \
    --path $TESTPATH/integration/test_processing.py
$ ./service-test worker --keep \
    -- path $TESTPATH/integration/test_processing.py:TestProcessing.test_200_checksum

Run tests matching a substring:

$ ./service-test worker --keep -k SUBSTRING

Recreate the database template, if the database has changed:

$ ./service-test worker --keep --build

If you prefer, you can also execute the commands above from within the container:

$ docker compose -f docker-compose.testing.yml up -d
$ docker compose -f docker-compose.testing.yml exec test_worker bash

    # run tests
    > service-test [--path PATH] [PARAMETER...]

    # rebuild database template
    > db-rebuild --no-template -d production collecting_society_test_template

    # exit container
    > exit

$ docker compose -f docker-compose.testing.yml down

The rendered HTML output of the coverage can be accessed via:

xdg-open volumes/shared/cover_worker/index.html

Pyramid¶

Run all pyramid tests once:

$ ./service-test web

Run all pyramid tests and keep the container running for the next test run:

$ ./service-test web --keep

Stop the container afterwards:

$ ./service-test --down

Note

The following commands will use the --keep flag by default. It will highly speed up the execution time, if you run the tests more than once.

You can append the normal pytest parameters:

$ ./service-test web --keep [--path PATH] [PARAMETER]

Run all tests quietly, drop into pdb on errors, don’t suppress output:
```
$ ./service-test web --keep --quiet --pdb -s
```

Run a specific set of tests:

$ ./service-test web --keep --path PATH[/FILE[:CLASS[.METHOD]]]

For example:

$ ./service-test web --keep \
    --path code/portal_web/portal_web/tests/unit
$ ./service-test web --keep \
    --path code/portal_web/portal_web/tests/unit/resources.py
$ ./service-test web --keep \
    --path code/portal_web/portal_web/tests/unit/resources.py:TestResources
$ ./service-test web --keep \
    --path code/portal_web/portal_web/tests/unit/resources.py:TestResources.test_add_child

Run a specific type of tests:

$ ./service-test web --keep --path (unit|functional|integration)

For example:

$ ./service-test web --keep --path unit
$ ./service-test web --keep --path functional
$ ./service-test web --keep --path integration

Run tests matching a substring:

$ ./service-test web --keep -k SUBSTRING

Recreate the database template, if the database has changed:

$ ./service-test web --keep --build

If you prefer, you can also execute the commands above from within the container:

$ docker compose -f docker-compose.testing.yml up -d
$ docker compose -f docker-compose.testing.yml exec test_web bash

    # run tests
    > service-test [--path PATH] [PARAMETER...]

    # rebuild database template
    > db-rebuild --no-template -d production collecting_society_test_template

    # exit container
    > exit

$ docker compose -f docker-compose.testing.yml down

Note

In the testing environment, the webgui and webapi services run both on the web service as deployment needs to be coordinated and controlled by pytest.

The rendered HTML output of the coverage can be accessed via:

xdg-open volumes/shared/cover_web/index.html

The screenshots of the selenium integration tests can be found in the folder:

volumes/shared/tmp/screenshots/

Linting¶

Lint the code for the scripts in this repository:

python -m flake8 scripts

Lint the code for application repositories via container:

docker compose exec SERVICE service-lint
docker compose exec SERVICE service-lint all
docker compose exec SERVICE service-lint --path /some/path/to/lint

Note

The code is also linted in the service-test script.

Demodata¶

The datasets are imported via a custom data import module using proteus with a trytond backend (not via XMLRPC). The most important files and folders are:

`volumes/shared/data/main.py`	Main function
`volumes/shared/data/datasets/__init__.py`	Definition of Dataset(s) classes
`volumes/shared/data/datasets/MODEL.py`	Dataset generation script for tryton model
`volumes/shared/data/csv/MODEL.csv`	CSV file for tryton model
`volumes/shared/data/csv/MODEL.py`	Script to generate the CSV file for tryton model

A minimal working dataset consists of two attributes:

#!/usr/bin/env python
DEPENDS = []            # A list of other datasets to be build first
generate(reclimit=0):   # The function to generate the datasets
    pass

Note

The dataset production is a special stage tag to separate the provision, which is neccessary for technical reasons from pure demodata.

Rebuild¶

In the development and staging environment, the demodata is created automatically during the setup of the database. If you need to rebuild the database, just use your prefered method:

via db-rebuild script:
```
$ ./db-rebuild
```
via db-rebuild CLI command on a running container:
```
$ docker compose exec erpserver db-rebuild
```

via db-rebuild CLI command on a new container:

$ docker compose run --rm erpserver db-rebuild

via db-rebuild CLI command inside the erpserver container:
```
> db-rebuild
```

The generation script will output some useful information during the run:

Configuration of the run
Name of the dataset
Description of the dataset
Models created/deleted/copied/updated and Wizards executed
Duration of the generation

Update¶

If you want to change a certain dataset for a model without constantly generating the demo data from scratch, this workflow is highly recommended:

Apply the changes to datasets/MODEL.py.
Test your changes by generating the MODEL dataset using the db-rebuild CLI command:
```
$ docker compose run --rm erpserver bash
> db-rebuild -d MODEL
```
While there are errors, fix them and retest using the --cache flag:
```
> db-rebuild -d MODEL --cache
```
Retest the whole generation:
```
> db-rebuild
```
Commit the changes.

If you want to change several datasets, you can prepare a template for the most time consuming master dataset and start the data generation from it with the -e/--exclude flag:

> db-rebuild --no-template -d production collecting_society_template
> db-rebuild -e production -d <DATASET>

You can also prepare a template for any dataset and copy it for later use:

> db-rebuild --no-template -d production collecting_society_artist
> db-copy --force collecting_society_artist collecting_society_template
> db-rebuild -e artist -d <DATASET>

Create¶

If you want to create a new dataset, you can use this template and take a look at the other datasets to see, how it works:

#!/usr/bin/env python
# -*- coding: UTF-8 -*-
# For copyright and license terms, see COPYRIGHT.rst (top level of repository)
# Repository: https://github.com/C3S/collecting_society_docker

"""
Create the <MODEL>s
"""

from proteus import Model

DEPENDS = [
    '<DATASET>',
]


def generate(reclimit=0):

    # constants

    # models

    # wizards

    # entries

    # content

    # create <MODEL>s

Note

All datasets/*.py files are registered automatically as new datasets on each run.

Documentation¶

The documentation is built with Sphinx and integrates the documentation of all collecting society applications. It contains both the *.rst files (e.g. README.rst) of the application repositories, as well as the python code api generated via autoapi.

The build process runs on a special documentation service container, because for autoapi the python modules need to be imported. To create the image for the container on the first built, use the --build flag of the docs-build script:

$ ./docs-build --build

To build the documentation afterwards, you can then just use:

$ ./docs-build

If you edit the documentation and need to build it more than once, you can use the --keep flag, to keep the container running and use the command successively:

$ ./docs-build --keep

To stop and remove the container, when you have finished, enter

$ ./docs-build --down

If you did not change any *.py files, you can use the --no-autoapi flag to omit the autoapi step and speed up the build:

$ ./docs-build --keep --no-autoapi

If you prefer, you can also execute the commands above from within the container:

$ docker compose -f docker-compose.documentation.yml up -d
$ docker compose -f docker-compose.documentation.yml exec documentation bash

    # build documentation via script
    > docs-build

    # build with autoapi omitted
    > docs-build --no-autoapi

    # exit container
    > exit

$ docker compose -f docker-compose.documentation.yml down

The main source files can be found in the ./volumes/shared/docs/source/ folder.

Warning

Don’t edit the *.rst files in the subfolders, because those are symlinked or generated by autoapi.

Once built, the docs can be viewed (from outside the container) like this:

$ xdg-open docs/index.html

See also

Sphinx rst Markup

Upgrade¶

These instructions perform a full upgrade of the

docker debian base image

docker upstream images

pip packages

repositories

Note

These instructions suppose, that the last current version is running properly. If not, just rebuild all containers at once after pinnings were removed and fix dependend containers as neccessary.

Note

Follow the instructions consecutivly.

Preperations¶

Create feature branch
./project checkout feature-upgrade
Update Dockerfile
vi ./services/build/Dockerfile
- Update image version tags
  
  base_production: Current debian stable version
  
  database: Current postgres package major version
  
  browser: Latest selenium/standalone-firefox tag
  
  webserver: Latest nginx-proxy tag
  
  sao_compiled: Latest node tag
- Update debian package names/versions (search for apt-get in Dockerfile)
- Remove version pinnings of pip packages:
  ./services/build/pip_unpin.sh

Browser¶

Update docker image

docker compose -f docker-compose.testing.yml build test_browser

Run service
docker compose -f docker-compose.testing.yml up test_browser
- Fix startup errors
- Update .env(.example) files and document changes in README.rst

Webserver¶

Update docker image
docker compose build webserver
Run service
docker compose up webserver
- Fix startup errors
- Update .env(.example) files and document changes in README.rst

Run healthcheck

docker compose up webserver -d
docker compose exec webserver bash
> /shared/healthcheck/webserver; echo $?

Fix healthcheck script errors

Database¶

Update docker image
docker compose build database
Update environment
docker compose up database
- Fix startup errors
- Update .env(.example) files and document changes in README.rst

Run healthcheck

docker compose up database -d
docker compose exec database bash
> /shared/healthcheck/database; echo $?

Fix healthcheck script errors

Fingerprint¶

Update docker image
- Update Dockerfile
  vi ./services/build/Dockerfile
  
  Update version of tokyocabinet
  
  Update version of tokyotyrant
  
  Update version of echoprint-codegen
- Build docker image
  docker compose build fingerprint
  
  Fix build errors
- Start docker container
  docker compose run --rm fingerprint bash
  
  Fix startup errors
Run service
docker compose up fingerprint
- Fix cli script
- Fix runtime errors
- Fix echoprint-server errors

Run healthcheck

docker compose up fingerprint -d
docker compose exec fingerprint bash
> /shared/healthcheck/fingerprint; echo $?

Fix healthcheck script errors

Erpserver¶

Update docker image

Build docker image
docker compose build erpserver
Fix build errors
Start docker container
docker compose run --rm --service-ports erpserver bash
Fix startup errors

Update tryton version
- Change TRYTON_VERSION in .env[.example]
  vi .env vi .env.example
- Change version in tryton.cfg of tryton modules
  vi ./code/collecting_society/tryton.cfg
- Change proteus version in install_requires of collecting_society_worker setup.py
  vi ./code/collecting_society_worker/setup.py
- Change tryton version in install_requires of portal_web setup.py
  vi ./code/portal_web/setup.py
- Ensure clean src repositories und update them
  ./project status ./project update
- Diff erpserver.py and update it accordingly
  diff services/deploy/erpserver.py volumes/shared/src/tryton/trytond/bin/trytond vi services/deploy/erpserver.py
Update collecting_society tryton module
- Delete *.pyc files (always on bad magic number import error)
  sudo find ./volumes/shared/src -name \*.pyc -delete
- Run service
  docker compose run --rm erpserver bash
  
  Install pip packages
  > pip install -r /shared/config/pip/erpserver.pip
  
  Fix cli script
  
  Fix pip dependencies
  
  Upgrade module files (see Release Announcements in News and Migration Forum)
  
  Run tests
  
  Run trytond tests
  > python -m unittest trytond.modules.collecting_society.tests.test_module
  
  Fix trytond tests
  
  Run scenario tests
  > python -m unittest trytond.modules.collecting_society.tests.test_scenario
  
  Fix scenario tests
  
  Run all tests
  > service-test
  
  Run demodata import
  
  Import until dataset upgrade
  > db-rebuild -d upgrade --reclimit 1 --pdb
  
  Fix demodata import script: ./volumes/shared/data/main.py
  
  Fix proteus stats: ./volumes/shared/data/main.py:ProteusStats
  
  Fix upgrade problems
  
  Import from dataset upgrade until dataset production
  > db-copy --force collecting_society collecting_society_template > db-rebuild -e upgrade -d production --reclimit 1 --pdb
  
  Fix datasets (see <module>/tests/scenario_*.txt for examples)
  
  Import from dataset production all remaining datasets
  > db-copy --force collecting_society collecting_society_template > db-rebuild -e production --reclimit 1 --pdb
  
  Fix datasets (see <module>/tests/scenario_*.txt for examples)
  
  Test a full database rebuild
  > db-rebuild
  
  Run linter
  > service-lint
  
  Fix linter errors
  
  Ensure, that the server is running
  > service-deploy
  
  Test Tryton Web Client interface

Run healthcheck

docker compose up erpserver -d
docker compose exec erpserver bash
> /shared/healthcheck/erpserver; echo $?

Fix healthcheck script errors

Connect tryton client
- Install the new tryton client and connect it to the server (see Tryton Installation and Tryton Usage)
- Test the client and operations (list, show, write, etc.)
  
  Open each menu item
  
  Open entries
  
  Check tabs of entries
  
  Fix methods

Worker¶

Update docker images
- Build docker image
  docker compose build worker
  
  Fix build errors
- Start docker container
  docker compose run --rm worker bash
  
  Fix startup errors
Update collecting_society_worker files
docker compose run --rm worker bash
- Install pip packages
  > pip install -r /shared/config/pip/worker.pip
  
  Fix pip dependencies
- Run service
  > service-deploy
  
  Fix echoprint-server errors
  
  Fix runtime errors
- Run tests
  > service-test
  
  Fix runtime errors
  
  Fix test errors
- Run linter
  > service-lint
  
  Fix linter errors

Run healthcheck

docker compose up worker -d
docker compose exec worker bash
> /shared/healthcheck/worker; echo $?

Fix healthcheck script errors

Web¶

Update docker images

Build docker images
docker compose build webgui webapi
Fix build errors
Run docker container
docker compose run --rm --service-ports webgui bash
Fix startup errors

Update *_web files
- Run service
  docker compose run --rm --service-ports webgui bash
  
  Install pip packages
  > pip install -r /shared/config/pip/webgui.pip
  
  Fix pip dependencies
  
  Upgrade *_web files (see Upgrading, Changes, Change History)
  
  Start service
  > service-deploy
  
  Fix changes of tryton api, especially in
  
  portal_web/__init__.py
  
  portal_web/models/base.py:Tbd
  
  portal_web/config.py
  
  Fix changes of pyramid api
  
  Test frontend / backend (login, show, write, etc)
  
  Fix deprecation warnings
- Run tests (on host)
  Remove docker volumes
  docker compose -f docker-compose.testing.yml down docker volume rm \ collecting_society_test_echoprint_data \ collecting_society_test_postgresql_data \ collecting_society_test_trytond_files
  
  Build test images
  ./service-test --build
  
  Fix build errors
  
  Run tests (see Pyramid Tests)
  ./service-test web --keep --path unit ./service-test web --keep --path functional ./service-test web --keep --path integration ./service-test web --keep
  
  Fix test wrapper errors: ./code/portal_web/tests/base.py
  
  Fix startup errors
  
  Fix test errors
  
  View screenshots of integration tests: ./tests/screenshots
  
  Fix linter errors

Documentation¶

Update docker image

docker compose -f docker-compose.documentation.yml build documentation

Build documentation

docker compose -f docker-compose.documentation.yml run --rm documentation bash
> docs-build

Fix build errors
Check documentation build
xdg-open docs/index.html

Wrap-up¶

1. Run all tests with build flag (on host)

./service-test --build --keep

Update versions
- Update Dockerfile
  List upgraded pip versions
  docker compose -f docker-compose.documentation.yml \ run --rm documentation pip freeze
  
  Add version pinnings of pip packages
  ./services/build/pip_pin.sh
  
  Ensure matching versions of selenium pip package and docker image version tag
  
  Build docker images
  docker compose build --no-cache docker compose -f docker-compose.testing.yml build docker compose -f docker-compose.documentation.yml build
  
  Run all tests with build flag (on host)
  ./service-test --build
- Update project.yml
  vi ./project.yml
  
  Update reference repository tags of task checkout repos of pinned pip packages for reference
  
  Ensure CHECKOUT_REFERENCES=1 in .env
  
  Update repositories
  ./project update
  
  Fix checkout errors
  
  Fix unwanted diff lines
  
  Ensure all envars are present in *.env and *.env.example
- Update links in Reference section of wiki
Update documentation
- Enhance the documentation, where possible
- Build documentation
  ./docs-build --keep --no-autoapi
- Check all changes
  xdg-open docs/index.html

Push changes to remote feature branch

./project status
./project diff -v
./project commit "updates <SERVICE>: <SOURCEVERSION> -> <TARGETVERSION>"
./project push

Request tests by other team members
Merge the feature branch
- Merge changes of development branch into the feature branch first
  ./project checkout development ./project pull ./project merge --no-delete --no-push feature-upgrade
  
  Fix merge conflicts and commit
  ./project commit "fixes merge conflicts"
  
  Run tests
  ./project checkout feature-upgrade ./service-test --build --keep
  
  Commit and push
  ./project commit "fixes stuff" ./project push
- Merge feature branch into development branch
  ./project checkout feature-upgrade ./project merge
  
  Wait for the result of the Jenkins build
  
  Fix the Jenkins job environment, if neccessary
  
  Note the upgrade procedure for the news entry instructions
  
  Fix/commit/push the code, until it runs successfully
  
  Check the documentation built
- Consider to merge the development branch into the staging branch
  ./project promote development
  
  Fix the Jenkins job environment, if neccessary
  
  Adjust all .env and config files manually, if neccessary (e.g. on tryton version change)
  
  Check output of db-update for migration errors
- Update the online documentation for development and staging
Notify team members
- Write a news entry with instructions how to upgrade and a link to the diff of the changes, note that the changes to Dockerfile contain all relevant version changes
  https://github.com/c3s/collecting_society_docker/compare/<OLD_COMMIT_ID>..<NEW_COMMIT_ID>

Problems¶

Docker¶

Couldn’t connect to Docker daemon

Docker compose cannot start container <id> port has already been allocated

If docker fails to start and you get messages like this: “Couldn’t connect to Docker daemon at http+unix://var/run/docker.sock […]” or “docker compose cannot start container <docker id> port has already been allocated”

Check if the docker service is started:
```
$ sudo systemctl start docker
```
Check if any user of docker is member of group docker:
```
$ login
$ groups | grep docker
```

Tryton¶

Bad Fingerprint

If the Tryton client already connected the tryton-container, the fingerprint check could restrict the login with the message: Bad Fingerprint!

That means the fingerprint of the server certificate changed. In production use, the Bad fingerprint alert is a sign that someone could try to fish your login credentials with another server responding your client. Ask the server administrator if the certificate has changed.

Close the Tryton client. Check the problematic host entry in ~/.config/tryton/7.0/known_hosts. Add a new fingerprint provided by the server administrator or simply remove the whole file, if the setup is not in production use:

rm ~/.config/tryton/7.0/known_hosts

Incompatible Server Version

If the tryton client shows an “incompatible server version” error on login try:

rm ~/.config/tryton/7.0/known_hosts

License¶

For infos on copyright and licenses, see ./COPYRIGHT.rst.