Installation best practises
It is highly recommended to set up an encrypted connection when accessing the Tracardi API. One way to do this is to encode the traffic as it is sent to the load balancer, but leave the internal cluster traffic decoded. This allows for secure communication with the API while still allowing for efficient communication within the cluster. If you prefer to have encoded traffic throughout the entire cluster, you will need to prepare HTTPS versions of the Tracardi Docker images. This will ensure that all traffic, both external and internal, is encrypted for added security. It is important to carefully consider the security measures in place when working with sensitive data, and setting up an encrypted connection to the Tracardi API is an important step in protecting this information. See the documentation on how to do it.
Separation of track server and GUI API
Best practice is to have a setup of 2 Tracardi API clusters. One for GUI and one for event collection.
The event collecting cluster, the one available on the Internet, should be configured with environment
EXPOSE_GUI_API set to
no. This way the publicly available API will have only
/track endpoint available.
All other endpoints that are required by the GUI will be disabled. This cluster will be used for event collection from
the website or other sources available on the Internet.
Another cluster, the one available in the internal network, or available on the Internet but restricted to certain set
of IPs, should have the environment variable
EXPOSE_GUI_API set to
yes. This cluster will be used by GUI to control
Scaling refers to the process of increasing or decreasing the capacity of a system to meet the demands of incoming traffic. In the context of Tracardi, scaling may involve increasing the number of servers or instances (tracardi-api) in order to handle a larger volume of event tracking requests.
Tracardi is a stateless system, meaning that it does not rely on storing data locally in order to function. This makes it easier to scale, as new instances of Tracardi can be added or removed without affecting the overall functioning of the system. We recommend using Kubernetes, a container orchestration platform, to manage the scaling and maintenance of the Tracardi cluster. Tracardi does not require long-lasting sticky sessions, so it can be easily killed and re-launched without disrupting the cluster.