You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Since 4.20, a few things changed when configuring a data storage location. Instead of a single place for everything, you may configure multiple storages and choose between those per Dataverse.
Support for defining the storage locations is already present via env vars from the ConfigMap
By default, dataverse.files.storage-driver-id points to file. We will create and use one with id local to be more human readable.
We should provide an opt-out default file storage based on /data.
This is mimicking the current behaviour, but enables flexibility when no local storage should be available (production instances) but deserved for testing, docker-compose usage etc.
When dataverse.files.storage-driver-id=local, we configure this as our local storage. When someone configures some other storage as default (like an S3 storage), it will most likely either have a different id or at least will overwrite the necessary configuration bits.
By default, dataverse.files.directory=/data will continue to drop any temp data (uploads, ...) at /data/temp. (The temp dir is created by Dataverse).
The docs need to reflect the changes:
Update example ConfigMap (S3 storage options!)
Fix strange inclusion line numbers of configuration help page.
Add a link to the storage explanaition in upstream docs, as those aren't well described in the JVM options overview.
On storage page, describe the /data/temp thing and it's corresponding JVM option.
Also on storage page, describe options how to configure multiple storages plus link to and from configuration page. (So basically this sounds like a new section about remote storage and an addition about local storage)
The text was updated successfully, but these errors were encountered:
Since 4.20, a few things changed when configuring a data storage location. Instead of a single place for everything, you may configure multiple storages and choose between those per Dataverse.
dataverse.files.storage-driver-id
points tofile
. We will create and use one with idlocal
to be more human readable./data
.dataverse.files.storage-driver-id=local
, we configure this as our local storage. When someone configures some other storage as default (like an S3 storage), it will most likely either have a different id or at least will overwrite the necessary configuration bits.dataverse.files.directory=/data
will continue to drop any temp data (uploads, ...) at/data/temp
. (Thetemp
dir is created by Dataverse).The docs need to reflect the changes:
The text was updated successfully, but these errors were encountered: