Skip to main content
This page covers what your Redshift administrator sets up. The steps you take inside Inflection are in Import from Data Warehouse and Export to Data Warehouse.

​
Prerequisites

  • A Redshift superuser, or a user who can create users and grant privileges on the database.
  • Access to the cluster’s security group in the AWS console, to allow Inflection’s IP addresses.
  • For import, source tables or views that meet the source table requirements in Import from Data Warehouse.

​
Connection Fields

Import connection Export connection uses the same fields plus a Schema where Inflection writes the export tables. Find the hostname and port under Endpoint in the cluster details. In the example below, the hostname is redshift-cluster-1.cmjhsmtjaqgq.ap-south-1.redshift.amazonaws.com and the port is 5439. Redshift cluster endpoint details showing hostname and port

​
Network Access

Required. Inflection connects from static IP addresses, and Redshift rejects the connection until they are allowed.
  1. In the AWS console, open the cluster’s security group and add an inbound rule for each Inflection IP address on the cluster port. The import wizard’s Connect step lists them under Allow these IPs first, and the export setup shows the same list.
  2. In the wizard, select I have allowed these IPs. Test connection is enabled only after you do.
Connect step for Amazon Redshift with the Allow these IPs first list and the I have allowed these IPs checkbox

​
Prepare Redshift for Import

​
Create the User

Create a user for Inflection with a password. See CREATE USER in the Redshift documentation.

​
Grant Access to Your Data

In Step 2 of the import wizard, the Warehouse access chip checks read access on your source schemas and write access on the write location. If anything is missing, it shows the exact statements to run, built from your database, user, schemas, and write schema. Click Copy statements, run them, then click Re-check. The generated statements have this shape. Run the generated version, not this one.
On Redshift the write location is always a schema in the database you connect to. One connection reads from one database. To read from a second database, create a second connection.
If you use RA3 nodes with cross-database queries, keep the write schema in the same database as the sources. Cross-database writes are not supported.

​
Create Views Yourself (Optional)

When you choose We draft a view for you in the wizard, Inflection writes the CREATE OR REPLACE VIEW statement for you. You can either click Create it for me, which runs it with the write grant above, or click Copy SQL, run it yourself, and click Verify View. You can also build your own view ahead of time, grant SELECT on it to inflection_reader, and pick it as a source like any table.

​
Prepare Redshift for Export

  1. Allow Inflection’s IP addresses as described under Network Access.
  2. Create a user and a schema for the export tables, and grant the user access to the schema.
  3. After the export connection is created in Inflection, open it, click Connection Details, and copy the Table Creation SQL. Run it in Redshift. It creates the two export tables with the structure Inflection writes and grants INSERT, UPDATE, and DELETE on them. Sample SQL. Do not use this directly. Please refer to the SQL shown on inflection UI for your connection
Once the tables exist, tell your Inflection team. They validate the setup and activate the sync.

​
Troubleshooting

Confirm the inbound rules in the cluster security group include every Inflection IP address on the cluster port, and that the Host is the endpoint hostname without a protocol prefix or a trailing :5439.
The user needs CREATE on the database to create the write schema, or the schema must already exist with ALL granted to the user. Run the generated statements as a superuser, then click Re-check.
Intervals between 24 and 72 hours keep data volumes manageable on Redshift. Contact your Inflection team to change the interval.