Object Storage
This document outlines the configuration process for adding and managing system object storages within Practicus AI.
Overview
The System Object Storages section enables you to integrate external object storage services (e.g., AWS S3) for managing and accessing data objects, such as model files and other resources.
Adding a New Object Storage
To add a new object storage, click the Add System Object Storage button and fill in the following fields:
- Key: (Required) A unique name for the object storage used for administrative purposes.
- Endpoint URL: (Required) The URL of the object storage service. For AWS S3, use the format
https://s3.{aws_region}.amazonaws.com
where{aws_region}
is the region (e.g.,us-east-1
). - Bucket Name: (Required) The name of the object storage bucket where data will be stored. Models will be stored under the path
[endpoint_url]/[bucket_name]/models/[model_host_key]/[model_key]/[version]/[model_files]
. - Prefix: (Optional) A prefix path to use before accessing objects. For example, if the prefix is
some/prefix
, objects will be accessed usinghttps://[endpoint]/[bucket]/some/prefix/object.data
. - Access Key ID: (Required) The access key for the object storage service.
- Secret Access Key: (Required) The secret key for the object storage service.
Actions
- Save: Saves the new object storage configuration.
- Save and Add Another: Saves the configuration and opens a new form to add another object storage.
- Save and Continue Editing: Saves the configuration but keeps the form open for further adjustments.