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

# How to connect

> How to connect the Redshift data source to Adriel: allowlist the egress IP, enter cluster credentials, select a table, and verify data.

Redshift is Amazon's cloud data warehouse. Connecting it binds a customer-selected table (or view) from a provisioned cluster or Serverless workgroup as a dashboard data source, with fields discovered directly from the table schema so warehouse rollups, customer-360 tables, and attribution models appear alongside marketing connectors.

## Before you connect

The following are required:

* An Amazon Redshift cluster (provisioned) or Serverless workgroup that is reachable over the network.
* The cluster's security group configured to allow inbound traffic from Adriel's egress address, `52.79.160.224`, on the cluster port — Redshift's default is `5439`. Without this rule the connection times out before any query runs.
* A dedicated Redshift user with read-only `SELECT` on the target table only. The connector issues read queries exclusively, so a read-only role is sufficient.
* Connection details gathered ahead of time: host name (or, for Serverless, the workgroup endpoint), port, database name, username, and password.

To find the host and database in AWS, open **Amazon Redshift** → **Clusters** → the cluster → the **Properties** tab → **Connection details**. The host name is the endpoint string ending in `.amazonaws.com`; the database name is the final path segment of the connection string (for example, the `dev` in `...redshift-serverless.amazonaws.com:5439/dev`).

## Connect Redshift

<Steps>
  <Step title="Open the connector">
    On the **Data Sources** page, search for **Redshift** and select it.
  </Step>

  <Step title="Enter the connection details">
    Fill in the connection form for the target cluster:

    * **Name**: any identifiable name for the data source.
    * **Host**: the Redshift host name or Serverless workgroup endpoint from the AWS console.
    * **Port**: the cluster port. Redshift's default is `5439`.
    * **Database**: the database to fetch from.
    * **User**: the read-only user with `SELECT` on the target tables.
    * **Password**: the password for that user.
  </Step>

  <Step title="Select the table">
    Choose the table (or view) to connect. Multiple selections are supported, and each connected table becomes its own data source.
  </Step>

  <Step title="Submit">
    Click **Submit** to complete the connection. Confirm `52.79.160.224` is allowlisted if the cluster sits behind a firewall. The initial sync runs in the background and data can take up to one business day to appear.
  </Step>
</Steps>

## What gets imported

Each column of the selected table becomes a field, typed from its native Redshift type: numeric columns become metrics, date columns become the date field for widget-level date filtering through **Blend Data settings**, and text and boolean columns become breakdown fields. Every widget query is capped at 1,000 rows and that cap is fixed; to return fewer, wider rows, pre-aggregate on the Redshift side into a rolled-up table or view and bind that object instead.

For the full field type mapping, query model, caching, and limits, see the [Redshift data reference](/data-sources/o-z/redshift/data-reference).

## Troubleshooting

<AccordionGroup>
  <Accordion title="The connection times out or is refused">
    Adriel reaches the cluster through its database proxy, so the cluster's security group must allow inbound traffic from `52.79.160.224` on the cluster port (default `5439`). If that address is not permitted, or the host name does not resolve, the connection fails before any query runs. For Redshift Serverless, use the workgroup endpoint in place of a cluster endpoint.
  </Accordion>

  <Accordion title="Authentication fails">
    The connecting user either has the wrong credentials or lacks read access. Confirm the read-only user can `SELECT` from the target table, then re-enter the password when reconnecting.
  </Accordion>

  <Accordion title="Results are truncated at 1,000 rows">
    The 1,000-row cap is fixed and cannot be raised per data source. Pre-aggregate the data inside Redshift into a rolled-up table or view and bind that, so each group arrives as a single row.
  </Accordion>

  <Accordion title="An aggregated text or boolean value looks wrong">
    Text and boolean columns are not summed. When a group holds a single distinct value it is returned as-is; when a breakdown collapses several distinct values the connector falls back to a single representative value, which can look unexpected. Use text and boolean columns for grouping rather than aggregation, or pre-aggregate inside Redshift when a stable single value is required.
  </Accordion>
</AccordionGroup>

## Related

<CardGroup cols={2}>
  <Card title="Redshift data reference" href="/data-sources/o-z/redshift/data-reference">
    Field type mapping, query model, caching, and limits for the Redshift data source.
  </Card>

  <Card title="Redshift FAQs" href="/data-sources/o-z/redshift/faqs">
    Common questions and expected behaviors for the Redshift data source.
  </Card>

  <Card title="How to connect PostgreSQL" href="/data-sources/o-z/postgres/how-to-connect">
    The Postgres-family SQL connector Redshift most closely mirrors.
  </Card>

  <Card title="How to connect MySQL" href="/data-sources/g-n/mysql/how-to-connect">
    Connect a MySQL database as a data source.
  </Card>

  <Card title="How to connect Snowflake" href="/data-sources/o-z/snowflake/how-to-connect">
    Connect the alternative cloud data warehouse.
  </Card>
</CardGroup>