> ## Documentation Index
> Fetch the complete documentation index at: https://docs.domino.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Data

> Where your data lives when you work in Domino, and how to choose between Data Source connectors, Domino Datasets, NetApp Volumes, and External Data Volumes.

Domino separates your data from your code. Your code lives in Project files, which Domino versions on every change. Your data lives outside those files, so your commit history stays about code, and the same data can serve any Project that needs it. Domino gives you several ways to reach data, and the right one depends on where the data already lives and how you want to work with it.

You don't have to move data into Domino to use it. You can query databases, warehouses, and object stores where they already run, mount existing NFS or SMB file shares, plug in enterprise NetApp storage, or store data inside Domino when you want managed, shareable storage with built-in snapshots backed by the platform.

The Data section covers all of these methods, in the order they appear in the sidebar:

* **[Domino Datasets](/6.3/platform-capabilities/core-concepts/data/datasets)**: Managed filesystem storage inside Domino, shareable across Projects and versioned with snapshots. Every Project starts with its own Dataset, named after the Project, which is the fastest way to write data alongside your code without polluting your Project's file history.
* **[NetApp Volumes](/6.3/platform-capabilities/core-concepts/data/netapp-volumes)**: NetApp ONTAP enterprise storage mounted directly into Domino workloads. For organizations that run ONTAP, NetApp Volumes match what Datasets do and add enterprise-grade storage features. See [Datasets vs. NetApp Volumes](#datasets-vs-netapp-volumes) for how to choose.
* **[External Data Volumes (EDVs)](/6.3/platform-capabilities/core-concepts/data/external-volumes)**: Existing Network File System (NFS) or Server Message Block (SMB) file shares that an administrator registers with Domino and mounts into your executions.
* **[Access external data directly](/6.3/platform-capabilities/core-concepts/data/access-external-data-directly)**: Reach any data service from your code using the same libraries you'd use locally, when Domino doesn't have a managed connector for it.
* **[Data sharing and security](/6.3/platform-capabilities/core-concepts/data/sharing-and-security)**: How access control works for each method.
* **[Domino Data Sources](/6.3/platform-capabilities/core-concepts/data/data-source-connectors)**: Managed connections to popular external databases, warehouses, and object stores that you query from code, without hand-managing drivers or credentials.

[Project artifacts](/6.3/platform-capabilities/core-concepts/projects/git-based-projects#save-and-pull-artifacts) round out the picture: a version-controlled folder inside each Project, sized for small outputs such as charts and serialized models, and the one place where "data" is intentionally tied to a single Project's history.

### Data across the platform

Data connections in Domino are platform-level objects, not Project contents: the same Data Source, Dataset, NetApp Volume, or EDV can be [attached to every Project](/6.3/platform-capabilities/core-concepts/projects) that needs it, with its own permissions, and each Project's executions mount or query the attached data. Project artifacts are the exception that lives inside a Project.

Connected data then serves the whole lifecycle, not just development. [Workspaces](/6.3/platform-capabilities/core-concepts/workspaces), [Jobs](/6.3/platform-capabilities/core-concepts/jobs), [Flows](/6.3/platform-capabilities/features/flows), and [on-demand compute clusters](/6.3/platform-capabilities/features/compute-clusters-hpc) read and write it during development and training. Published [Apps](/6.3/platform-capabilities/features/apps), [Agents](/6.3/platform-capabilities/features/agents), and [model APIs](/6.3/platform-capabilities/features/model-deployment) query it at serving time. [Model Monitoring](/6.3/platform-capabilities/features/monitoring) connects to your data sources on its own to compare live predictions against ground truth, and its drift and quality alerts drive retraining. [Governance](/6.3/platform-capabilities/features/governance) closes the accountability loop: Dataset files can serve as evidence in governance bundles, and Data Source activity is recorded in the [Audit Trail](/6.3/platform-capabilities/features/governance/audit-trail).

```mermaid placement="top-right" theme={null}
flowchart LR
    ext[("Your external data:<br/>databases, warehouses,<br/>file shares")]
    subgraph domino ["The Domino platform"]
        subgraph shared ["Platform-level data, shared across Projects"]
            ds["Data Source<br/>connectors"]
            edv["External Data Volumes"]
            nv["NetApp Volumes"]
            dsets[("Domino Datasets:<br/>Domino-managed storage")]
        end
        proj["Projects:<br/>attach the data they need<br/>(artifacts live here)"]
        exec["Executions:<br/>Workspaces, Jobs, Apps,<br/>Agents, model APIs"]
    end
    ext --> ds
    ext --> edv
    ext --> nv
    ds --> proj
    edv -->|attached to| proj
    nv -->|attached to| proj
    dsets --> proj
    proj -->|mounted and queried by| exec
```

## Choose a method

Two defaults cover most work:

* **Your data lives in an external store**: Use a Data Source connector. Your administrator can pre-configure the connection details centrally, so you add only your credentials: no guessing at hostnames or port numbers. Credentials are stored securely, and the connection is reusable in any Project. If Domino has no connector for your store, [connect directly](/6.3/platform-capabilities/core-concepts/data/access-external-data-directly) with the same library you would use locally.
* **You want managed storage inside Domino**: Use a Domino Dataset, or a NetApp Volume if your organization runs ONTAP. Both mount into executions as filesystem paths, both support snapshots for reproducibility, and both are shareable across Projects. See [Datasets vs. NetApp Volumes](#datasets-vs-netapp-volumes) below.

The other methods fit specific situations: use an External Data Volume when IT already maintains an NFS or SMB file share you need, and Project artifacts for small outputs that belong with your code.

### Datasets vs. NetApp Volumes

Datasets and NetApp Volumes cover the same ground. Anything Datasets do, NetApp Volumes generally do at least as well and often better, but they aren't strictly "Domino-managed" the way Datasets are: an administrator needs a NetApp ONTAP account configured and connected first. Once that's done, users work with NetApp Volumes the same way they work with Datasets, from the same **Data** menu.

Use **Datasets** when:

* You want storage that works out of the box, with no infrastructure prerequisites.
* You are creating a data artifact scoped to a Project (Datasets originate inside a Project, though you can then share them anywhere).
* You want to branch your data: you can create a new Dataset from a Dataset snapshot, where a NetApp Volume snapshot can't be cloned into a new NetApp Volume.

Use **NetApp Volumes** when:

* Your organization already runs, or is willing to set up, NetApp ONTAP.
* You need multi-terabyte, high-performance shared storage with enterprise backup, snapshot retention, and disaster recovery.
* You want snapshots that don't consume additional storage capacity and near-instant snapshot creation.
* You want data to be a global entity in Domino rather than a Project-owned artifact.

## Compare the methods

| Method                                                                                                            | Best for                                                                     | Where the data lives           | Typical size                              | Access control                                                                                                                                         |
| ----------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------- | ------------------------------ | ----------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ |
| [Domino Datasets](/6.3/platform-capabilities/core-concepts/data/datasets)                                         | Versioned training data and pipeline outputs, out of the box                 | Domino-managed storage         | Gigabytes to a few terabytes              | [Dataset roles](/6.3/platform-capabilities/core-concepts/data/sharing-and-security/share-datasets)                                                     |
| [NetApp Volumes](/6.3/platform-capabilities/core-concepts/data/netapp-volumes)                                    | Multi-terabyte shared storage with enterprise backup and efficient snapshots | Your NetApp ONTAP filesystems  | Multi-terabyte and beyond                 | [NetApp Volume roles](/6.3/platform-capabilities/core-concepts/data/netapp-volumes/create-netapp-volumes#add-users-or-organizations-to-netapp-volumes) |
| [External Data Volumes](/6.3/platform-capabilities/core-concepts/data/external-volumes)                           | NFS or SMB file shares IT already maintains                                  | Your NFS or SMB storage        | Whatever your NFS or SMB storage provides | [Assigned users and groups](/6.3/platform-capabilities/core-concepts/data/sharing-and-security/share-edv)                                              |
| [Data Source connectors](/6.3/platform-capabilities/core-concepts/data/data-source-connectors)                    | Querying existing databases, warehouses, and object stores                   | Your external data service     | What you pull into the executor           | [Users, groups, or a service account](/6.3/platform-capabilities/core-concepts/data/sharing-and-security/share-data-sources)                           |
| [Project artifacts](/6.3/platform-capabilities/core-concepts/projects/git-based-projects#save-and-pull-artifacts) | Small outputs that belong with code                                          | The Project's artifact storage | About 10 GB in total                      | [Project collaborators](/6.3/platform-capabilities/core-concepts/projects/collaborate-on-projects#collaborator-permissions)                            |

### Sizing and cost

Domino sets no size limit on Datasets, NetApp Volumes, or External Data Volumes: the ceiling comes from whatever storage backs them, so check that provider's documentation for the figure that applies to you. For NetApp Volumes it depends on the ONTAP volume type, and a single traditional FlexVol supports up to 300 TiB. Project files, where artifacts live, are the exception, and Domino applies its own configurable limits there. See [Files, artifacts, and results](/6.3/platform-capabilities/core-concepts/projects/files-artifacts-and-results).

What grows is cost, not a Domino quota. Storage cost tracks the bytes you keep, and every Dataset snapshot duplicates the data it captures, so delete what you no longer need. NetApp's redirect-on-write snapshots consume no extra space to start, which keeps the cost of versioning low. When cost matters more than filesystem access, keep bulk data in object storage such as Amazon S3 and reach it through a Data Source connector.

Administrators can cap what any one person stores. See [Dataset quotas and limits](/6.3/platform-capabilities/core-concepts/data/datasets/manage-large-data#dataset-quotas-and-limits).

## Data reproducibility

When your work must be repeatable, use snapshots: immutable, versioned copies of a Dataset or NetApp Volume that your code can target by version number or tag. Data Source connectors and EDVs read live data, so Domino can't guarantee that the same query returns the same result later. If the external system supports its own snapshots or time travel, you can use those features from your code, or copy query results into a Dataset or NetApp Volume and snapshot them.

## Share your data securely

Data in Domino is reusable across Projects with full control over who can access it. Datasets, NetApp Volumes, Data Sources, and EDVs each carry their own permissions, so sharing a Project does not automatically expose its data. Dataset and NetApp Volume sharing is zero-copy: every Project mounts the same files rather than receiving a copy, so a read-only shared Dataset or NetApp Volume works as a controlled way to distribute a curated data product across teams. See [Data sharing and security](/6.3/platform-capabilities/core-concepts/data/sharing-and-security) for how to grant and restrict access for each method.

## Work with data

* [Datasets](/6.3/platform-capabilities/core-concepts/data/datasets): Create, use, and snapshot Domino Datasets.
* [NetApp Volumes](/6.3/platform-capabilities/core-concepts/data/netapp-volumes): Mount ONTAP volumes into Domino workloads.
* [External Data Volumes](/6.3/platform-capabilities/core-concepts/data/external-volumes): Mount existing network file shares.
* [Access external data directly](/6.3/platform-capabilities/core-concepts/data/access-external-data-directly): Reach any other data service with its native library.
* [Data sharing and security](/6.3/platform-capabilities/core-concepts/data/sharing-and-security): Control who can read and edit your data.
* [Domino Data Sources](/6.3/platform-capabilities/core-concepts/data/data-source-connectors): Connect to supported databases, warehouses, and object stores.


## Related topics

- [Data Planes](/6.3/platform-capabilities/core-concepts/data-plane.md)
- [Manage data](/6.3/admin/data-administration/index.md)
- [Get Data Source Audit Data](/6.3/api-reference/datasource/get-data-source-audit-data.md)
- [Use Data Sources](/6.3/platform-capabilities/core-concepts/data/data-source-connectors/use-data-sources.md)