Docker

Docker

Server and analytics processes work in separate docker containers. There are several ways to run them which are listed here.

Run via docker-compose

You can use docker-compose setup for more useful and declarative management of containers. For this you need docker-compose.yml file that contains services configuration and .env file that contains variables default values which are placed here:

• docker-compose.yml
• .env

These files should be placed in one folder.

How to get HASTIC_API_KEY is described here.

export HASTIC_API_KEY=<your_grafana_api_key>
docker-compose up -d

Run manually

docker run -d \
  --name hastic-analytics \
  -p 8002:8002 \
  hastic/analytics

docker run -d \
  --name hastic-server \
  -p 8000:8000 \
  -e HASTIC_API_KEY=<your_grafana_api_key> \
  -e ZMQ_CONNECTION_STRING=tcp://<analytics_ip>:8002\
  -v /some/host/path:/var/www/data \
  hastic/server

NOTE: if you use Grafana and Hastic on one host, use any host IP (e.g. 192.168.0.1) except for localhost (or 127.0.0.1) when Grafana os opened in browser. Otherwise Hastic will try to find Grafana inside Docker and fail.

ZMQ_CONNECTION_STRING is the string which is used for connection between server and analytics. For docker-compose by default the value tcp://analytics:8002 from .env file will be used. For manual setup use tcp://<analytics_ip>:8002 with WAN IP (for example 192.168.0.1).

In Docker, Hastic uses TCP connection. If you want to increase performance use IPC in production mode as described in Installation from source.