The recently merged PR #6300 added support for Elasticsearch, but it did not include a graphical user interface. This PR adds a separate service that includes Kibana, providing a GUI for managing and querying Elasticsearch.

Notes

1. Elasticsearch & Kibana v9 We’re using the latest version (v9) of Elasticsearch and Kibana. Version 9 does not support password-based authentication between Kibana and Elasticsearch. Instead, it requires a service account token.

   • This service includes a kibana-token-generator container that automatically generates the token and prints it in the logs.
   • Users must copy this token and set it as the value for the ELASTICSEARCH_SERVICEACCOUNTTOKEN environment variable.
   • Without this, Kibana will not be able to connect to Elasticsearch.

2. Memory Considerations Elasticsearch is resource-intensive by default. Many self-hosted Coolify users run on low-memory servers, which can cause the instance to become unresponsive.

   • To mitigate this, we set ES_JAVA_OPTS=-Xms512m -Xmx512m to limit JVM heap size to 512MB.

3. Default User The Elasticsearch service uses a default user named elastic. This username cannot be changed, so it’s hardcoded in the Docker Compose config.

4. Startup Time Elasticsearch + Kibana can take over 2 minutes to start fully. Please be patient during initial deployment.

Deployment Guide

If any of the above is unclear, here’s a video showing how to deploy and configure this service properly.

https://github.com/user-attachments/assets/b7a81062-0604-4394-9268-4d0fc8ea6fac

Bounty

/claim #6449

Docs PR

https://github.com/coollabsio/coolify-docs/pull/348