View a markdown version of this page

Querying system tables - Amazon Redshift
Example Anti-pattern vs. Recommended Queries

Amazon Redshift will no longer support the use of Python UDFs after June 30, 2026. We will start enforcing it in phases. For more information on the details of Python end of life and migration options, see the blog post that was published on June 30, 2025.

Querying system tables

If neither the driver metadata API nor SHOW commands meet your needs, you can query PostgreSQL catalog tables (pg_*), the information schema (information_schema), or Amazon Redshift system views such as SVV_TABLE_INFO directly.

This approach is useful for ad hoc exploration, debugging, or administrative tasks that you run from interactive SQL queries, for example in Amazon Redshift query editor v2 or psql. However, system catalog structures can change across Amazon Redshift releases. We don't recommend that you rely on direct system table queries as the metadata interface for production application integrations.

When you query system tables, follow these practices:

  • Specify database, schema, and table. For best performance, include the database name, schema name, and table name in your query filters. Narrowing the scope reduces the amount of metadata that Amazon Redshift needs to scan and return.

  • Use equality or LIKE filters. Use equality (=) or LIKE predicates in your WHERE clause to narrow results. Avoid unfiltered queries against large catalog views. They can be slow on clusters with many objects.

  • Avoid functions in filter predicates. Don't use functions such as LOWER() or string concatenation in your WHERE clause when you filter system table queries. These prevent efficient metadata lookups and can significantly degrade query performance.

Example Anti-pattern vs. Recommended Queries

-- Avoid:
SELECT * FROM svv_all_columns;

-- Recommended:
SELECT * FROM svv_all_columns WHERE database_name = 'dev' and schema_name = 'public' AND table_name = 'sales';
-- Avoid:
SELECT * FROM svv_all_columns WHERE LOWER(schema_name) = 'public';

-- Recommended:
SELECT * FROM svv_all_columns WHERE schema_name = 'public' AND table_name LIKE 'sales%';
-- Avoid:
SELECT * FROM svv_all_columns WHERE database_name || '.' || schema_name || '.' || table_name = 'dev.public.sales';

-- Recommended:
SELECT * FROM svv_all_columns WHERE database_name = 'dev' and schema_name = 'public' AND table_name = 'sales';
-- Avoid:
SELECT * FROM svv_all_columns WHERE CONCAT(database_name, '.', schema_name) = 'dev.public';

-- Recommended:
SELECT * FROM svv_all_columns WHERE database_name = 'dev' and schema_name = 'public';