For users who prefer full control over their data or have existing storage infrastructures, External Storage provides the flexibility to integrate cloud or on-premise storage solutions directly into the VESSL platform. This allows users to connect their preferred storage without migrating data into VESSL.
volume://{storage-name}/{volume-name}
. VESSL supports AWS S3 and Google Cloud Storage as external cloud storage options. Follow the steps below to integrate cloud storage:
Integrate cloud credentials
Integrating AWS S3 or Google Cloud Storage is straightforward. Navigate to Organization settings > Integrations and add your credentials.
Integrate external storage
After adding credentials, integrate your external storage using the bucket path. Click Test connection
to validate the credentials, and then click Integrate
.
Create a new volume
Create a volume in the external storage for use in workloads.
Import the volume to run
When creating a run in the web console, create a directory (e.g., /dataset
) and import the volume from External Storage as follows:
Alternatively, you can manually specify the volume in the YAML editor:
Export a workload volume to storage
To export your workload’s data to External Storage, create a new directory (e.g., /output
), and export the data as follows:
Alternatively, you can manually specify the export in the YAML editor:
Locate the exported volume
Navigate to Run > Files to find he file tree of the executed run.
Follow the export volume link to access the data in External Storage.
VESSL supports NFS (Network File System) and host path volumes for users who maintain on-premise storage solutions. You can integrate and use data from on-premise storage by following these steps:
Integrate external storage
Integrate
button.Mount the volume to a run
NFS volumes can be directly mounted to runs, reducing initialization time and making it ideal for large datasets.
Alternatively, you can manually specify the mount in the YAML editor:
Similar to VESSL Storage, you can tag external storage volumes for easy search and retrieval. Tags are supported across both cloud and on-premise storage, ensuring a consistent experience.