Lakehouse Dynamic Table Development Quick Start Guide

Overview

Dynamic Tables are the core vehicle for incremental computation in Singdata Lakehouse. You simply declaratively define the query logic, and the system automatically detects upstream data changes and incrementally refreshes results, eliminating the need to manually write complex incremental ETL logic. This guide is organized by development workflow to help you quickly master dynamic table creation, refresh, and monitoring methods.

Create Dynamic Table -- Declaratively define incremental computation logic
Manual Refresh -- Use REFRESH to trigger computation immediately
View Refresh History -- Monitor refresh status and mode
View Dynamic Table Structure -- Use DESC to view definition
Drop Dynamic Table -- Use DROP DYNAMIC TABLE to clean up

SQL Commands Covered

Command	Purpose	Use Case
`CREATE DYNAMIC TABLE ... AS SELECT`	Create dynamic table	Declaratively define incremental computation logic
`REFRESH DYNAMIC TABLE`	Manual refresh trigger	Compute latest results immediately
`SHOW DYNAMIC TABLE REFRESH HISTORY`	View refresh history	Monitor refresh status
`DESC DYNAMIC TABLE`	View dynamic table structure	View definition and configuration
`DROP DYNAMIC TABLE`	Drop dynamic table	Clean up dynamic tables no longer needed

Prerequisites

The following examples use simulated orders table orders_dt and products table products_dt:

-- Create orders table CREATE TABLE IF NOT EXISTS orders_dt ( order_id INT, product_id INT, quantity INT, order_date DATE ); -- Create products table CREATE TABLE IF NOT EXISTS products_dt ( product_id INT, product_name STRING, category STRING ); -- Insert test data INSERT INTO products_dt VALUES (1, 'iPhone 15', 'Phone'), (2, 'MacBook Pro', 'Laptop'), (3, 'AirPods', 'Audio'); INSERT INTO orders_dt VALUES (1, 1, 2, '2024-06-01'), (2, 2, 1, '2024-06-01'), (3, 1, 1, '2024-06-02');

Create Dynamic Table

Use CREATE DYNAMIC TABLE to declaratively define computation logic. The system automatically handles incremental computation.

-- Create a dynamic table summarizing by product category CREATE DYNAMIC TABLE dt_category_sales REFRESH INTERVAL 10 MINUTE VCLUSTER default AS SELECT p.category, COUNT(*) as order_count, SUM(o.quantity) as total_quantity FROM orders_dt o JOIN products_dt p ON o.product_id = p.product_id GROUP BY p.category;

Parameter Descriptions:

REFRESH INTERVAL 10 MINUTE: Automatically refresh every 10 minutes.
VCLUSTER default: Specifies the compute cluster used for the refresh.

⚠️ Note: After creation, it is recommended to immediately execute REFRESH once to reset the refresh time baseline.

Manual Refresh

Use REFRESH DYNAMIC TABLE to trigger computation immediately without waiting for scheduled refresh.

-- Manually trigger refresh REFRESH DYNAMIC TABLE dt_category_sales;

Result Verification:

SELECT * FROM dt_category_sales ORDER BY category;

category	order_count	total_quantity
Laptop	1	1
Phone	2	3

💡 Tip: After inserting new data into the source table, execute REFRESH to see the incremental computation results.

View Refresh History

Use SHOW DYNAMIC TABLE REFRESH HISTORY to monitor refresh status and determine whether it was an incremental or full refresh.

-- View the last 5 refresh records SHOW DYNAMIC TABLE REFRESH HISTORY WHERE name = 'dt_category_sales' LIMIT 5;

Key Field Descriptions:

state: Refresh status (SUCCEED / FAILED / RUNNING)
refresh_mode: Refresh mode (INCREMENTAL / FULL)
refresh_trigger: Trigger method (MANUAL / SYSTEM_SCHEDULED)

View Dynamic Table Structure

Use DESC DYNAMIC TABLE to view the dynamic table's definition and configuration.

-- View dynamic table structure DESC DYNAMIC TABLE dt_category_sales;

Returned Information:

Table structure (column names, types)
Refresh interval
VCluster used
Underlying SQL definition

Incremental Computation Verification

Insert new data into the source table to verify the dynamic table's incremental refresh capability.

-- Insert a new order INSERT INTO orders_dt VALUES (4, 3, 5, '2024-06-03'); -- Manual refresh REFRESH DYNAMIC TABLE dt_category_sales; -- View updated results SELECT * FROM dt_category_sales ORDER BY category;

Result Explanation:

category	order_count	total_quantity
Audio	1	5
Laptop	1	1
Phone	2	3

⚠️ Note: The Audio category changed from 0 to 1, demonstrating that incremental computation correctly captured the new data.

Drop Dynamic Table

Use DROP DYNAMIC TABLE to drop a dynamic table. Note that you must use DROP DYNAMIC TABLE, not DROP TABLE.

-- Drop dynamic table DROP DYNAMIC TABLE dt_category_sales;

💡 Tip: Dropped dynamic tables can be recovered via UNDROP TABLE within the Time Travel retention period.

Clean Up Test Data

After completing dynamic table verification, it is recommended to clean up test tables:

-- Drop test tables DROP TABLE IF EXISTS orders_dt; DROP TABLE IF EXISTS products_dt;

💡 Tip: Lakehouse supports UNDROP TABLE, allowing recovery of accidentally dropped tables within the retention period.

Notes

VCluster Selection: Use GENERAL type VClusters (such as default), not ANALYTICS type.
Refresh Frequency: Higher frequency leads to higher CRU consumption. Use 1 DAY for T+1 scenarios, 1 HOUR for hourly.
Non-deterministic Functions: Avoid using non-deterministic functions like CURRENT_TIMESTAMP(), RAND() in Dynamic Tables.
DROP Syntax: You must use DROP DYNAMIC TABLE; using DROP TABLE will produce an error.
Incremental Limitations: Some complex queries (such as outer JOINs with non-equi conditions) may not execute incrementally; the system will automatically fall back to full refresh.