I'm using the elk stack in docker and I am looking for a way to configure the default columns displayed in the log stream. Right now those default columns are timestamp, event.dataset and message.
I can change them from the setting menu, but I haven't been able to reproduce this from the kibana config file.
unfortunately there's no mechanism to change the columns via the configuration file at the moment. What makes using the settings UI impractical for you?
I am running my whole stack in a docker-compoe file and my main issue was about loosing Kibana settings when doing a docker-compose down.
I am only using the elastic stack to parse a single application log file for now so I don't have that many needs.
Anyway to solve that issue I have made a volume of elasticsearch data folder (as Kibana settings are saved there).
But logs are also saved there and initially I wanted to avoid this as I wanted elastic/filebeat to start from 0 and reparse the totally of my log file. Now it's on tail mode in order to avoid duplicate.
Thanks for providing the insights. I can empathize - automatically restoring Kibana settings after a redeployment is not trivial. Maybe I can help by providing some insights into how the log source settings are stored:
Upon saving the settings in the Logs UI a space-specific saved object is stored in the Kibana system index. You should be able to query such saved objects using the saved object _find API as in
curl -u "${USERNAME}:${PASSWORD}" -X GET '${KIBANA_URL}/api/saved_objects/_find?type=infrastructure-ui-source' -H 'kbn-xsrf: true'
and restore them using the _bulk_create API after container startup. A one-shot sidecar container is a common pattern for that.
In the meantime I have created https://github.com/elastic/kibana/issues/61773 to put more convenient export/import mechanisms on the roadmap. Please feel free to chime in with your use-case details there to provide some weight to the request.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.