The OpenSearch connector provides two types of storage for document content (or files).
File system
This connector enables files to be stored on a file system considered local by the JVM (local, NFS, etc.). The directory used can be configured:
file.dir=/opt/FlowerDocs/Files/
Amazon S3
The Amazon S3 connector stores document content in an S3 bucket. To use this connector, the following configuration is required:
core.services.file.dao=s3
s3.region=<region AWS>Paramétrage du client Amazon S3
| Property | Default value | Description |
|---|---|---|
| s3.max.connections | 100 | Maximum number of open HTTP connections |
| s3.max.error.retry | 2 | Maximum number of retries for replayable requests (error 5xx) |
| s3.socket.timeout | 100000 | Waiting time (in ms) for data to be transferred |
By default, the file containing the access and secret key pair must be located in the ${USER_HOME}/.aws/credentials folder, and the default profile used.
Another key file or profile can be used using the following parameters:
s3.profile=<nom du profil>
s3.configFilePath=<chemin du répertoire>/<fichier contenant les clés>AWS instance roles can be used by adding the following property:
s3.instanceProfile=trueA single bucket can be used for all FlowerDocs scopes with the following configuration:
s3.bucket.scoped=false
s3.bucketName=<nom du bucket>