Pre-requisites

  • The chosen database needs to exist in the cluster.
  • The chosen schema needs to exist in the cluster.
  • A user with access to create a new table and insert records to it in the chosen database and schema.

Setting up a write-only Redshift user

The following provides an example for how to set ups. user with write-only access to a specific table using SQL commands.

Create a user

# Assign a user name and password. Use a username and password of your choice.
CREATE USER validio_user WITH password 'securepassword'

# Create a group for the user. Use a group name of your choice.
CREATE GROUP validio_group;

# Add the newly created user to the created group.
ALTER GROUP validio_group ADD USER validio_user;

Grant user write only access to the configured table

# Grant access for the group to user schema containing the table.
GRANT USAGE ON SCHEMA my_schema TO GROUP validio_group;

# Grant write only access for the group to the table in the schema.
GRANT INSERT ON my_schema.my_table TO validio_group;

Redshift destination configuration parameters

⚠️

Validio currently supports standard unquoted identifiers for the following fields:

  • Database name
  • Schema
  • Table name

Please let us know if you have a use-case that requires support of another type of identifier.

Field Required Description
Name Identifier for the connector. Used when setting up pipelines.
Host Endpoint of the Redshift cluster database. e.g "test-redshift-cluster-1.ynypdanx10uk.eu-north-1.redshift.amazonaws.com"
Port Port number of the database.
Username Username of a Redshift user that can create the table in the given database and schema and then write records to it.
Password Password of the specified Redshift user.
Database name Name of the Redshift database. Must be a supported identifier.
Schema Name of the schema containing the table to write to. This can be left empty if the table resides in the default public schema. Must be a supported identifier.
Table name Name of the table from which to write the data. Validio will automatically attempt to create the table if it does not exist. Must be a supported identifier.

Destination format

Column name Type Description
time TIMESTAMP WITH TIME ZONE Time that the record was written.
is_anomaly BOOLEAN Indicate whether the datapoint is an anomaly.
applied_filters VARCHAR JSON list of applied filters to the datapoint and whether it was detected as an anomaly.
datapoint VARCHAR The data contained in the processed datapoint.