Skip to main content

CHANGES

Introduced or updated: v1.2.410

The CHANGES clause allows querying the change tracking metadata for a table within a defined time interval. Please note that the time interval must fall within the data retention period (defaulted to 24 hours). To define a time interval, use the AT keyword to specify a time point as the start of the interval, with the current time being applied as the default end of the interval. If you wish to specify a past time as the end of the interval, use the END keyword in conjunction with the AT keyword to set the interval.

alt text

Syntax

SELECT ...
FROM ...
CHANGES ( INFORMATION => { DEFAULT | APPEND_ONLY } )
AT ( { TIMESTAMP => <timestamp> |
OFFSET => <time_interval> |
SNAPSHOT => '<snapshot_id>' |
STREAM => <stream_name> } )

[ END ( { TIMESTAMP => <timestamp> |
OFFSET => <time_interval> |
SNAPSHOT => '<snapshot_id>' } ) ]
ParameterDescription
INFORMATIONSpecifies the type of change tracking metadata to be retrieved. Can be set to either DEFAULT or APPEND_ONLY. DEFAULT returns all DML changes, including inserts, updates, and deletes. When set to APPEND_ONLY, only appended rows are returned.
ATSpecifies the starting point of the time interval for querying change tracking metadata.
ENDOptional parameter specifying the end point of the time interval for querying change tracking metadata. If not provided, the current time is used as the default end point.
TIMESTAMPSpecifies a specific timestamp as the reference point for querying change tracking metadata.
OFFSETSpecifies a time interval in seconds relative to the current time as the reference point for querying change tracking metadata. It should be in the form of a negative integer, where the absolute value represents the time difference in seconds. For example, -3600 represents traveling back in time by 1 hour (3,600 seconds).
SNAPSHOTSpecifies a snapshot ID as the reference point for querying change tracking metadata.
STREAMSpecifies a stream name as the reference point for querying change tracking metadata.

Enabling Change Tracking

The CHANGES clause requires that the Fuse engine option change_tracking must be set to true on the table. For more information about the change_tracking option, see Fuse Engine Options.

Example:
-- Enable change tracking for table 't'
ALTER TABLE t SET OPTIONS(change_tracking = true);

Examples

This example demonstrates the use of the CHANGES clause, allowing for the tracking and querying of changes made to a table:

  1. Create a table to store user profile information and enable change tracking.
CREATE TABLE user_profiles (
user_id INT,
username VARCHAR(255),
bio TEXT
) change_tracking = true;


INSERT INTO user_profiles VALUES (1, 'john_doe', 'Software Engineer');
INSERT INTO user_profiles VALUES (2, 'jane_smith', 'Marketing Specialist');
  1. Create a stream to capture profile updates, then update an exiting profile and insert a new one.
CREATE STREAM profile_updates ON TABLE user_profiles APPEND_ONLY = TRUE;


UPDATE user_profiles SET bio = 'Data Scientist' WHERE user_id = 1;
INSERT INTO user_profiles VALUES (3, 'alex_wong', 'Data Analyst');
  1. Query changes in user profiles by the stream.
-- Return all changes in user profiles captured in the stream
SELECT *
FROM user_profiles
CHANGES (INFORMATION => DEFAULT)
AT (STREAM => profile_updates);

┌───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
│ user_id │ username │ bio │ change$action │ change$row_id │ change$is_update │
├─────────────────┼──────────────────┼───────────────────┼──────────────────┼────────────────────────────────────────┼──────────────────┤
1 │ john_doe │ Data Scientist │ INSERT69cffb02264144c384d56f7b6cedee41000000 │ true
3 │ alex_wong │ Data Analyst │ INSERT59f315c8655c49eab35ba1959e269430000000 │ false
1 │ john_doe │ Software Engineer │ DELETE69cffb02264144c384d56f7b6cedee41000000 │ true
└───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘

-- Return appended rows in user profiles captured in the stream
SELECT *
FROM user_profiles
CHANGES (INFORMATION => APPEND_ONLY)
AT (STREAM => profile_updates);

┌───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
│ user_id │ username │ bio │ change$action │ change$is_update │ change$row_id │
├─────────────────┼──────────────────┼──────────────────┼───────────────┼──────────────────┼────────────────────────────────────────┤
3 │ alex_wong │ Data Analyst │ INSERTfalse59f315c8655c49eab35ba1959e269430000000 │
└───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
  1. Query changes between a snapshot and a timestamp with both the AT and END keywords.
-- Step 6: Take a snapshot of the user profile data.
SELECT snapshot_id, timestamp
FROM FUSE_SNAPSHOT('default', 'user_profiles');

┌───────────────────────────────────────────────────────────────┐
│ snapshot_id │ timestamp
├──────────────────────────────────┼────────────────────────────┤
6a11c94433714970895edd38577ac8b0 │ 2024-04-10 02:51:39.422832
53dc4750af92423da91c50dcee547cfb │ 2024-04-10 02:51:39.399568
910af7424f764891b0c6fa60aa99fc3a │ 2024-04-10 02:50:14.522416
1225000916f44819a0d23178b2d0d1af │ 2024-04-10 02:50:14.500417
└───────────────────────────────────────────────────────────────┘

SELECT *
FROM user_profiles
CHANGES (INFORMATION => DEFAULT)
AT (SNAPSHOT => '1225000916f44819a0d23178b2d0d1af')
END (TIMESTAMP => '2024-04-10 02:51:39.399568'::TIMESTAMP);

┌──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
│ user_id │ username │ bio │ change$action │ change$row_id │ change$is_update │
├─────────────────┼──────────────────┼──────────────────────┼──────────────────┼────────────────────────────────────────┼──────────────────┤
1 │ john_doe │ Data Scientist │ INSERT69cffb02264144c384d56f7b6cedee41000000 │ true
1 │ john_doe │ Software Engineer │ DELETE69cffb02264144c384d56f7b6cedee41000000 │ true
2 │ jane_smith │ Marketing Specialist │ INSERT3db484ac18174223851dc9de22f6bfec000000 │ false
└──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
Did this page help you?
Yes
No
Explore Databend Cloud for FREE
Low-cost
Fast Analytics
Easy Data Ingestion
Elastic Scaling
Try it today