Skip to content

JupyterHub DataMount Spawner Configuration

Overview

This section provides a detailed explanation of the available configuration options for the JupyterHub DataMount Spawner.

Configuration Features

Templates

c.KubeSpawner.templates = ["b2drop", "aws", "s3", "webdav", "generic"]
  • Allows administrators to specify which mount templates should be visible to users.
  • The configured templates are passed to both the Jupyter Server and sidecar API. The API rejects mount requests for templates outside this allowlist, including requests made directly from notebook code.
  • If templates is unset or empty, the spawner uses uftp, b2drop, aws, s3, webdav, nfs, and generic.
  • List of currently supported templates can be seen here (class MountDialogComponent , attribute templates_all)

Logging Configuration

c.KubeSpawner.logging_config = {
  "stream": {
    "enabled": True,
    "level": 20,
    "formatter": "simple",
    "stream": "ext://sys.stdout",
  },
  "file": {
      "enabled": True,
      "level": 20,
      "filename": "/mnt/data_mounts/mount.log",
      "formatter": "simple_user", # simple_user, simple or json
      "when": "h",
      "interval": 1,
      "backupCount": 0,
      "encoding": None,
      "delay": False,
      "utc": False,
      "atTime": None,
      "errors": None,
  },
  "syslog": {
    "enabled": False,
    "level": 20,
    "formatter": "json",
    "address": ["ip", 5141],
    "facility": 1,
    "socktype": "ext://socket.SOCK_DGRAM",
  },
  "smtp": {
    "enabled": False,
    "level": 50,
    "formatter": "simple",
    "mailhost": "mailhost",
    "fromaddr": "smtpmail",
    "toaddrs": [],
    "subject": "SMTPHandler - Log",
    "secure": None,
    "timeout": 1,
  }
}
  • Configures Python’s logging handler to monitor the extension’s usage.
  • Supported logging handlers: stream, file, syslog, smtp. stream and file are enabled by default.
  • The log format and verbosity level can be customized as needed.

The file LogHandler is used as feedback for the user. Keep that in mind when editing its configuration.

Init Mounts

c.KubeSpawner.init_mounts = [
  {
    "path": "aws_pre",
    "options": {
      "displayName": "AWS (external)",
      "template": "aws",
      "config": {
        "remotepath": "myawsbucket",
        "type": "s3",
        "provider": "AWS",
        "access_key_id": "<ACCESS_KEY>",
        "secret_access_key": "<SECRET_KEY>",
        "region": "eu-north-1"
      }
    }
  }
]
  • Enables automatic mounting of external storage when JupyterLab starts.
  • Mounted storage is available to the user without manual intervention.
  • The mount is marked administrator-managed: its configuration is redacted by the API, and the API rejects attempts to remove it.
  • Secrets are not exposed by the mount-list API response, but administrators should still protect the Hub configuration and the sidecar's configuration volume.

By leveraging these configuration options, administrators can fine-tune the JupyterHub DataMount Spawner for their specific deployment needs.

Remember Mounts

To allow users to remember mounts that persist after JupyterLab restarts, each user must have a user-specific persistent storage. Their mount configurations will be stored in plain text at a place that an administrator has to configure. It must be configured via KubeSpawner.data_mount_config.

c.KubeSpawner.data_mount_config = """
c.DataMount.remember_enabled = True # default is False
c.DataMount.remember_file = "/home/jovyan/.jupyter/datamounts/mounts.json" # update if your persistent storage is at $HOME/work
c.DataMount.remember_default = False # configure whether the checkbox should be enabled by default. Default: False
"""

Enable NFS

By default NFS is disabled, since there's no authentication in plain NFS setups.

c.KubeSpawner.enable_nfs_mounts = True # default is false
c.KubeSpawner.blocked_nfs_mounts = ["192.168.0.0/24"] # User's won't be able to mount any Server within that CIDR. Can be an empty list, or multiple CIDR strings in a list.

NFS server addresses must be IP literals. The spawner's default block list includes common private, loopback, and link-local IPv4 ranges; configure the list for your network and threat model.