Getting started light-eventuate-4j
This tutorial will show you how to start light-eventuate-4j services with docker-compose. It is very convenient for developers who want to do an integration test on his/her laptop or desktop. For the official test environment and production, it is not recommended to use docker-compose or docker. These services should be installed in a data center or virtual machines to form a cluster.
Kafka needs a network IP address for ADVERTISED_HOST_NAME, and it cannot be localhost or 127.0.0.1 as localhost within docker container means the docker container itself. It must be an address on the host that is accessible within docker container. It should be the hostname of your computer or the real network IP address of your computer that runs docker.
There is a script in light-docker that can try to find the hostname and export it as DOCKER_HOST_IP, but it is not reliable as the hostname might map to 127.0.0.1 on some computers. The best way to do that would manually find out the IP address on your computer and export it.
Manually export DOCKER_HOST_IP
On Mac/Linux, issue the following command on a terminal
You will see a list of network interfaces in the output. Depending on your network setup, the interface name might be en0 on Mac and enp4s0 on Linux. You need to find the IP start with
192.XXX.XXX.XXX Or 10.XXX.XXX.XXX
Once you get the IP address, copy it to the clipboard. The next step is to export it to an environment variable DOCKER_HOST_IP.
Now you can start eventuate docker-compose from the same terminal you export the DOCKER_HOST_IP.
Automatically export DOCKER_HOST_IP on Mac
The above approach is the most reliable way to get the right IP address, and on my Mac Book Pro, it is the only way as the computer name is not mapped to a real IP address somehow.
The above approach works until you reboot your computer. If you want to setup a static IP for your Mac, then you can follow this Mac Permanent IP to set it up. It ensures that you can start the eventuate docker-compose at any terminal and you don’t need to restart it when you move from one WIFI hotspot to another WIFI hotspot.
Automatically export DOCKER_HOST_IP on Linux Desktop
At home, I am working on a Linux desktop with a fixed IP address allocated by the router.
If you are using a desktop which has a static IP address, then you can put the export statement into .bashrc under your home directory. It can eliminate the manual step and ensure that you can run the docker-compose from any terminal.
Here is a section of my .bashrc file on my desktop.
export PATH=$PATH:/home/steve/tool/apache-maven-3.5.0/bin export M2_HOME=/home/steve/tool/apache-maven-3.5.0 export M2=%M2_HOME%\bin export DOCKER_HOST_IP=192.168.1.120
Note that the above manual export and script will only work on the terminal window with the export and you have to start the docker-compose with the exact terminal in order to work as the environment variable only associates to that particular terminal instance.
Right after you have updated .bashrc, you can run the following command and then run the docker-compose from the same terminal. The next time you restart your computer, you can run the docker-compose in any terminal.
cd ~ source .bashrc
This approach only works with a desktop as it has a static IP. If you are using a laptop, most likely you will have to use the first option to export the IP manually or find a way to create an alias interface with a fixed IP address.
Start eventuate services
Now you have used a script or manually exported DOCKER_HOST_IP. The next step is to start eventuate services with a docker-compose file in the light-docker repository.
cd ~/networknt git clone https://github.com/networknt/light-docker.git cd light-docker
Now let’s start the docker-compose-integration-test.yml from the light-docker folder. Before we run the compose, we need to export the DOCKER_HOST_IP on this terminal unless you have updated .bashrc file.
After double checking the DOCKER_HOST_IP environment variable, you can issue the following command to start Kafka, Zookeeper, and Mysql altogether.
cd ~/networknt/light-docker docker-compose -f docker-compose-integration-test.yml up
Depending on how fast your computer is and if you have all the docker images cached locally, it might take about 30 seconds to 5 minutes to get everything started.
Start CDC server for eventuate
** The following section is only for CDC server developers. If you are creating applications based on light-eventuate-4j, light-tram-4j, and light-saga-4j you don’t need the steps below. **
Eventuate CDC is a server that monitors the events table in the database and sends the events to Kafka for other services to subscribe to. It is part of the docker-compose-integration-test.yml already so you don’t need to start it separately.
However, if you are a developer who is working on the eventuate CDC server, then you might want to comment out the eventuate-cdcserver from docker-compose-integration-test.yml and start eventuate CDC server by yourself.
cd ~/networknt/light-docker docker-compose -f docker-compose-cdcserver-for-eventuate.yml up
Another way to start CDC server is to start from the source code build. In this case, we need to compile light-eventuate-4j repo. Unless you are doing an update for the CDC server, you should use the above docker-compose to start CDC server for eventuate.
cd ~/networknt git clone https://github.com/networknt/light-eventuate-4j.git cd light-eventuate-4j mvn clean install -DskipTests cd eventuate-cdc-server java -jar target/eventuate-cdc-server.jar
Now we have the entire infrastructure up and running, and we can start some of the demo applications. The todo-list application is the easiest to get started with, and if you want an app with real business scenarios, check out account-management.