storage

⚠️ Disclaimers ⚠️ We do not take any responsibility on any data loss with the below recommendations!

By default, as adapted for developer environment, and not for production environment, where we want to persist all our data storage in Shanoir long time, Shanoir stores its content in Named Volumes, that are part of the local Docker Engine and Docker is taking care of them. More info on Named Volumes: Docker Documentation.

Therefore, see on the bottom of the docker-compose.yml, the declaration of all named volumes used:

• volumes:
• keycloak-database-data:
• rabbitmq-data:
• database-data:
• datasets-data:
• extra-data:
• studies-data:
• dcm4chee-ldap-data:
• dcm4chee-sldap-data:
• dcm4chee-database-data:
• dcm4chee-arc-wildfly-data:
• dcm4chee-arc-storage-data:
• solr-data:
• certificate-share-data:
• tmp:
• logs:

Please find below the named volumes necessary to adapt for long term, persistent production storage:

• keycloak-database-data: contains your user database
• database-data: contains all SQL data of all Shanoir microservices
• datasets-data: contains all NIfTI files generated during import
• extra-data: contains all extra files, attached to exams in shanoir
• studies-data: contains all extra files for studies, like DUAs and study protocols
• dcm4chee-database-data: contains all SQL data of the internal PACS of Shanoir
• dcm4chee-arc-storage-data: contains all DICOM files of the internal PACS of Shanoir
• certificate-share-data: could contain your certificate, but in production better store outside
• logs: contains all log files, important to persist as well

On your production environment you have two options to adapt the above storage to a long term solution:

1. You use local mount, so called Bind Mounts, and mount your local disk space Bind Mounts
2. You use the Named Volumes with volume drivers etc. and configure your corresponding solution

Take care of the final backup of your corresponding storage space for all data above.

In our config the volumes datasets-data, dcm4chee-arc-storage-data and studies-data are hosted directly on a NAS. The other volumes are local to the VM. In terms of volume, it is mainly the first two (datasets-data, dcm4chee-arc-storage-data) that are big (NIfTI and DICOM files). studies-data and extra-data contain the additional files attached to studies or datasets. We are in the order of 10 gigabytes for studies-data. extra-data is almost empty so we did not deport it.