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

# Zwiron Storage Connector

> Use Zwiron's built-in file storage as a pipeline source or destination — no external bucket needed.

**Zwiron Storage** is a built-in file storage system managed by Zwiron. Upload CSV, JSON, or Parquet files and use them directly as a pipeline source or destination — without setting up an S3 bucket, GCS bucket, or any external storage.

|                    |                                      |
| ------------------ | ------------------------------------ |
| **Source**         | ✅                                    |
| **Destination**    | ✅                                    |
| **Formats**        | CSV, JSON (JSONL), Parquet           |
| **Agent required** | No (runs on Zwiron's infrastructure) |

***

## When to use Zwiron Storage

* Ingest partner/vendor data drops (CSV files delivered via email or SFTP)
* Export pipeline output to a flat file for sharing or download
* Prototype pipelines before connecting real systems
* One-off data loads without setting up cloud storage credentials

***

## Connection setup

Zwiron Storage connections don't require credentials — the storage is managed by your organization. To create one:

1. Go to **Connections** → **New Connection**
2. Select **Zwiron Storage**
3. Choose a folder (or create a new one)
4. Save

***

## Using as a source

When Zwiron Storage is the source:

* Zwiron reads all files in the selected folder on each run
* Schema is inferred from the first file (or specified manually)
* Supported read options:

| Option          | Description                                                 |
| --------------- | ----------------------------------------------------------- |
| **Format**      | CSV, JSON, Parquet (auto-detected from extension)           |
| **Has header**  | For CSV — whether the first row contains column names       |
| **Delimiter**   | For CSV — comma (`,`), tab (`\t`), pipe (`\|`), etc.        |
| **Skip rows**   | Skip N rows at the start (for files with preamble metadata) |
| **Compression** | gzip, zstd (auto-detected)                                  |

### Scheduling with file sources

If files are replaced on a regular schedule (e.g. a partner drops a new `orders.csv` every morning), schedule the pipeline to run after the expected delivery time. Zwiron reads whatever is in the folder at that point.

***

## Using as a destination

When Zwiron Storage is the destination, Zwiron writes the sync output as files in the selected folder:

| Option           | Description                                                  |
| ---------------- | ------------------------------------------------------------ |
| **Format**       | CSV, JSON, Parquet                                           |
| **Partition by** | Write one file per table, or partition by date/size          |
| **Overwrite**    | Replace files on each run, or append with a timestamp suffix |

Output files appear immediately in the **Files** page and can be previewed or downloaded.

***

## Accessing your files

All files in Zwiron Storage are accessible from **Files** in the main navigation. You can:

* Browse folders
* Preview the first rows of any file
* Download files directly
* Upload new files manually

See [File Storage](/platform/file-storage) for the full file management UI.

***

## Limits

| Limit                    | Value                      |
| ------------------------ | -------------------------- |
| Max file size            | 10 GB per file             |
| Storage per organization | Varies by plan             |
| Formats                  | CSV, JSON (JSONL), Parquet |

For larger volumes or multi-region requirements, use a dedicated cloud storage connector: [Amazon S3](/connectors/files/s3), [Google Cloud Storage](/connectors/files/gcs), or [Azure Blob Storage](/connectors/files/azureblob).