Skip to content

Docker container installation

The easiest way to run Tracardi is to run it as a docker container.

Dependencies

Tracardi needs the following elements to run:

  • Elasticsearch
  • Redis
  • Tracardi API
  • Tracardi GUI

You will need to start the docker containers for all the above services.

Start ElasticSearch database

Tracardi need elasticsearch as its backend. Please pull and run elasticsearch single node docker before you start Tracardi.

You can do it with this command.

docker run -p 9200:9200 -p 9300:9300 -e "discovery.type=single-node" docker.elastic.co/elasticsearch/elasticsearch:7.13.2

Note

Running one instance of elasticsaerch is not a production solution. For production purposes, it is necessary to run the elasticearch cluster. You can also read here how to connect Tracardi to an elasticsearch cluster

Redis

You will need a redis instance as well.

Start it with:

docker run -p 6379:6379 redis

Start Tracardi API

Note

The following instalation description use the latest tracardi container version. If you would like to install stable version of the system, what we strongly recommend, please add to tracard/tracardi-api a tag with version, e.g tracardi/tracardi-api:0.7.1. The same applies to tracardi/tracardi-gui. Keep the version of API and GUI the same.

Run Tracardi API

Now pull and run Tracardi backend.

docker run -p 8686:80 -e ELASTIC_HOST=http://<your-laptop-ip>:9200 -e REDIS_HOST=redis://<your-laptop-ip>:6379 tracardi/tracardi-api #(1)
  1. Replace with your local laptop IP. You can remove -e REDIS_HOST=redis://<your-laptop-ip>:6379 if you did not start redis. Without redis some system features are unavailable.

Tracardi must connect to elasticsearch. To do that you have to set ELASTIC_HOST variable to reference your laptop's or server IP.

Waiting for application startup

Notice that when type http://localhost:9200 you try to connect to Elastic on localhost. This means that you're connecting to the docker itself as localhost means local in docker. Obviously elastic is not there, so Tracardi will never connect. Pass external ip for elastic. This may be your laptop IP if you are running Tracardi locally.

For more troubleshooting solutions go to Troubleshooting

Connecting Tracardi to ElasticSearch via SSL

If you have an elasticsearch instance and you would like to connect to it via HTTPS this is the command you may find useful.

docker run -p 8686:80 -e ELASTIC_HOST=https://user:password@<your-laptop-ip>:9200 -e ELASTIC_VERIFY_CERTS=no -e REDIS_HOST=redis://<your-laptop-ip>:6379 tracardi/tracardi-api #(1)
  1. If you want a certain version of docker image add version tag, e.g. tracardi/tracardi-api:0.7.1

ELASTIC_VERIFY_CERTS set to No

Notice that the above command does not verify SSL certificates. If you would like certificates to be validated set ELASTIC_VERIFY_CERTS to yes.

Start Tracardi GUI

Pull and run Tracardi Graphical User Interface.

docker run -p 8787:80 -e API_URL=//127.0.0.1:8686 tracardi/tracardi-gui #(1)
  1. If you want a certain version of docker image add version tag, e.g. tracardi/tracardi-gui:0.7.1

Note

Notice that the API_URL is set to localhost. If the API is located somewhere else replace localhost with the ip pointing to API.

Import worker

To run an import worker you will need the tracardi/worker to do the background importing of the data. Run the following command to start the worker. 

docker run -e REDIS_HOST=redis://<redis-ip>:6379 -d tracardi/worker #(1)
  1. If you want a certain version of docker image add version tag, e.g. tracardi/worker:0.7.1

Note

Replace the <redis-ip> with the IP of the redis instance. Keep the version of worker the same as tracard/tracardi-api

Start Tracardi Documentation

Pull and run Tracardi Documentation. This is the documentation you are reading right now

docker run -p 8585:8585 tracardi/tracardi-docs

Tracardi Graphical User Interface

Visit http://127.0.0.1:8787 and follow the instructions to set up Tracardi.

System Documentation

For the local copy of this documentation visit http://127.0.0.1:8585. The documentation docker must be started.

API Documentation

For API documentation visit http://127.0.0.1:8686/docs