Skip to main content

Views

A view is a logical table that encapsulates business logic and domain-specific semantics. While most compute engines support views natively, each engine stores view metadata in proprietary formats, creating interoperability challenges across different platforms. Paimon views abstracting engine-specific query dialects and establishing unified metadata standards. View metadata could enable centralized view management that facilitates cross-engine sharing and reduces maintenance complexity in heterogeneous computing environments.

Catalog support​

View metadata is persisted only when the catalog implementation supports it:

  • Hive metastore catalog – view metadata is stored together with table metadata inside the metastore warehouse.
  • REST catalog – view metadata is kept in the REST backend and exposed through the catalog API.
  • JDBC catalog – view metadata is stored in the catalog database in the paimon_views metadata table. The table is created automatically when the JDBC catalog is initialized.

File-system catalogs do not currently support views because they lack persistent metadata storage.

Representation structure​

FieldTypeDescription
querystringCanonical SQL SELECT statement that defines the view.
dialectstringSQL dialect identifier (for example, spark or flink).

Multiple representations can be stored for the same version so that different engines can access the view using their native dialect.

JDBC catalog notes​

The JDBC catalog stores views in a dedicated paimon_views table that is created on first initialization. A few things are worth knowing when running on top of an existing JDBC catalog:

  • Required permissions on upgrade. Upgrading to a Paimon release with view support requires CREATE TABLE permission on the catalog database the first time the catalog is opened, so that the paimon_views table can be created. Operators who tightened privileges to CRUD-only after the initial deployment should either restore CREATE TABLE permission temporarily or create the paimon_views table manually beforehand.
  • Table and view share the same identifier namespace. A name cannot be used by both a table and a view in the same database. createTable, renameTable, createView and renameView all validate this invariant under the catalog lock; concurrent operations targeting the same identifier will see exactly one winner.
  • Single-process atomicity does not depend on lock.enabled. The JDBC catalog also keeps a per-JVM stripe lock keyed by (catalog key, database, object name), so the table-vs-view name uniqueness invariant holds within one JVM even when lock.enabled = false. Setting lock.enabled = true (with lock.type = jdbc) is still recommended for multi-process deployments because the stripe lock only serializes operations within the same JVM.
  • Database visibility. A database that contains only views (and no tables or properties) is reported by listDatabases and SHOW DATABASES. DROP DATABASE ... CASCADE removes both the tables and the views in that database; DROP DATABASE without CASCADE will reject databases that still hold any view.
  • Cross-database rename. renameView(from, to) and renameTable(from, to) raise an IllegalArgumentException (Database X does not exist.) when the target database is missing, matching the BadRequest semantics of the REST catalog.

Operations​

Create or replace view​

Use CREATE VIEW or CREATE OR REPLACE VIEW to register a view. Paimon assigns a UUID, writes the first metadata file, and records version 1.

The catalog stores the view definition. It does not resolve or validate referenced tables when the view is created, so missing or cross-database table references are checked by the compute engine when the view is queried.

CREATE VIEW sales_view AS
SELECT region, SUM(amount) AS total_amount
FROM sales
GROUP BY region;

Alter view dialect via procedure​

Paimon provides the sys.alter_view_dialect procedure so that engines can manage multiple SQL representations for the same view version.

-- Add a Flink dialect
CALL [catalog.]sys.alter_view_dialect('view_identifier', 'add', 'flink', 'SELECT ...');

-- Update the stored Flink dialect
CALL [catalog.]sys.alter_view_dialect('view_identifier', 'update', 'flink', 'SELECT ...');

-- Drop the Flink dialect representation
CALL [catalog.]sys.alter_view_dialect('view_identifier', 'drop', 'flink');

Spark example​

-- Add a Spark dialect
CALL sys.alter_view_dialect('view_identifier', 'add', 'spark', 'SELECT ...');

-- Update the Spark dialect
CALL sys.alter_view_dialect('view_identifier', 'update', 'spark', 'SELECT ...');

-- Drop the Spark dialect
CALL sys.alter_view_dialect('view_identifier', 'drop', 'spark');

Drop view​

DROP VIEW view_name;

See also​