> For the complete documentation index, see [llms.txt](https://docs.rootcause.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.rootcause.ai/more-details/applications/application-modules/buffer.md).

# Buffer

> Accumulate arriving rows in a staging dataset and emit them as one batch when flushed or a bound is hit.

Collects rows across runs and releases them as one batch.

`items` appends to the staging dataset and `flush` drains it. Whichever of `maxItems` and `maxAgeS` is reached first also flushes; a null bound is no bound. A non-empty `dedupeKey` turns the append into a merge on those columns. This is one of the two module types that keep state between runs, so it is durable only.

## At a glance

|                          |                        |
| ------------------------ | ---------------------- |
| Module type              | `Buffer`               |
| Category                 | Data Processing        |
| Default tier             | `durable`              |
| Runs on                  | `durable`              |
| Starts a flow            | no                     |
| Holds a body             | no                     |
| Keeps state between runs | yes                    |
| Available in             | every application type |

## Parameters

### `dedupeKey`

`string[]` · required · defaults to `[]`

Non-empty turns the append into a merge on these columns.

### `maxItems`

`number | null` · optional · defaults to `null`

Flush once this many rows are staged. Null is no bound.

### `maxAgeS`

`number | null` · optional · defaults to `null`

Flush once the oldest staged row is this old. Null is no bound.

## Inputs

| Port    | Kind        | Required | Description                                                                                                                           |
| ------- | ----------- | -------- | ------------------------------------------------------------------------------------------------------------------------------------- |
| `on`    | `event`     | no       | Implicit on every module that does not itself start a flow: wire an event here to make this module the start of what that event runs. |
| `items` | `dataframe` | no       | Rows to add to the batch; `batch` stays empty until a flush, but `count` still reports.                                               |
| `flush` | `event`     | no       | Releases the batch now; an event port, so a schedule's `event` out-port wires straight in.                                            |

## Outputs

| Port    | Kind             | Connects to                        | Description                                                                                     |
| ------- | ---------------- | ---------------------------------- | ----------------------------------------------------------------------------------------------- |
| `batch` | `dataframe`      | `dataframe`, `signal`              | Everything staged since the last flush.                                                         |
| `count` | `scalar<number>` | `scalar`, `json`, `text`, `signal` | Rows waiting, emitted on every append — not the size of the batch, which only a flush produces. |

***

[← All application modules](/more-details/applications/application-modules.md)