This repository harvest two setups. The base of these setups resides in the standard docker-compose.yml.
- docker-compose.yml This provides you with the backend components. There is a frontend application included which you can publish using a separate proxy (we tend to put a letsencrypt proxy in front).
- docker-compose.dev.yml Provides changes for a good frontend development setup.
- publishes the virtuoso instance on port 8890 so you can easily see what content is stored in the base triplestore
- provides a mock-login backend service so you don't need the ACM/IDM integration.
General information on running and maintaining an installation
You'll need a beefy machine, with at least 16 GB of ram.
First install git-lfs
(see https://github.com/git-lfs/git-lfs/wiki/Installation)
# Ensure git-lfs is enabled after installation
git lfs install
# Clone this repository
git clone https://github.com/lblod/app-verenigingen-loket.git
# Move into the directory
cd app-verenigingen-loket
To ease all typing for docker-compose
commands, start by creating the following files in the directory of the project.
A docker-compose.override.yml
file with following content:
version: '3.7'
And an .env
file with following content:
COMPOSE_FILE=docker-compose.yml:docker-compose.dev.yml:docker-compose.override.yml
The loket app is huge, and a lot of data is being intialized. We want to make sure you don't overload your machine too much doing this the first time. It's an optional step, if you trust your machine is powerful enough, you can move on. This step should only be done once. First start virtuoso and let it setup itself
docker-compose up virtuoso
Wait for the logs
HTTP/WebDAV server online at 8890
Server online at 1111 (pid 1)
Then run the migrations
drc up migrations
This will take a while. You may choose to monitor the migrations service in a separate terminal to and wait for the overview of all migrations to appear: docker-compose logs -f --tail=100 migrations
. When finished it should look similar to this:
[2023-04-07 20:13:15] INFO WEBrick 1.4.2
[2023-04-07 20:13:15] INFO ruby 2.5.1 (2018-03-29) [x86_64-linux]
== Sinatra (v1.4.8) has taken the stage on 80 for production with backup from WEBrick
[2023-04-07 20:13:15] INFO WEBrick::HTTPServer#start: pid=13 port=80
This should be your go-to way of starting the stack.
docker-compose up # or 'docker-compose up -d' if you want to run it in the background
Always double check the status of the migrations docker-compose logs -f --tail=100 migrations
Wait for everything to boot to ensure clean caches.
Probably the first thing you'll want to do, is see wether the app is running correctly. The fastest way forward is creating a docker-compose.override.yml
file next to the other docker-compose.yml
files, and add
# (...)
loket:
ports:
- 4205:80
This way, you can directly connect to a built version of the frontend on port 4205
. Note, you might have conflicts because the port is already busy.
you're free to change 4205
to whatever suits you.
Once the migrations have ran, you can start developing your application by connecting the ember frontend application to this backend. See https://github.com/lblod/frontend-verenigingen-loket for more information on development with the ember application.
docker-compose up
The stack is built starting from mu-project.
OpenAPI documentation can be generated using cl-resources-openapi-generator.
To make sure the app can share data, producers need to be set up. There is an intial sync, that is potentially very expensive, and must be started manually
Due to performance issues, related to the high usage, a separate triplestore (virtuoso) has been introduced to offload the publication of the data. This architectural change is currently under evaluation. The criteria for evaluation will be: the performance win vs the practical consequences of such change.
If deemed succesful, we might consider moving the remaining publication graphs to this triplestore too (mandatarissen and leidinggevenden).
As a consequence, producers using the separate triplestore, will also publish and host the json-diff files. Mainly to simplify the transition to a separate publication triple store (else we would need a separate mu-auth and deltanotifier). In essence, it takes over https://github.com/lblod/delta-producer-json-diff-file-publisher, although both can still be combined.
If files need to be shared over deltas (attachments, form-data, cached-files) you will need to set in a docker-compose.override.yml
#(...)
delta-producer-publication-graph-maintainer-submissions:
KEY: "foo-bar
This will needs to be set in the consuming stack too. See delta-producer-publication-graph-maintainer for more informmation on the implications.
Once installed, you may desire to upgrade your current setup to follow development of the main stack. The following example describes how to do this easily for both the demo setup, as well as for the dev setup.
For the dev setup, we assume you'll pull more often and thus will most likely clear the database separately:
# This assumes the .env file has been set. Cf. supra in the README.md
# Bring the application down
docker-compose down
# Pull in the changes
git pull origin master
# Launch the stack
docker-compose up
As with the initial setup, we wait for everything to boot to ensure clean caches. You may choose to monitor the migrations service in a separate terminal to and wait for the overview of all migrations to appear: docker-compose logs -f --tail=100 migrations
.
Once the migrations have ran, you can go on with your current setup.
At some times you may want to clean the database and make sure it's in a pristine state.
# This assumes the .env file has been set. Cf. supra in the README.md
# Bring down our current setup
docker-compose down
# Keep only required database files
rm -Rf data/db
git checkout data/db
# Bring the stack back up
docker-compose up
Notes:
- virtuoso can take a while to execute its first run, meanwhile the database is inaccessible. Make also sure to wait for the migrations to run.
- data from external sources need to be synced again.