Can my DB?

CREATE FUNCTION set_updated_at()
RETURNS trigger LANGUAGE plpgsql AS $$
BEGIN
  NEW.updated_at := NOW();
  RETURN NEW;
END;
$$;

CREATE TRIGGER trg_set_updated_at
BEFORE UPDATE ON orders
FOR EACH ROW
EXECUTE FUNCTION set_updated_at();

• Trigger function must be created separately and return type trigger.
• Supports FOR EACH ROW and FOR EACH STATEMENT granularity.
• INSTEAD OF triggers work on views to make them updatable.
• DDL event triggers (CREATE EVENT TRIGGER) fire on DDL statements like CREATE TABLE.
• Transition tables (NEW TABLE, OLD TABLE) are available for statement-level triggers (v10+).

CREATE TRIGGER trg_set_updated_at
BEFORE UPDATE ON orders
FOR EACH ROW
BEGIN
  SET NEW.updated_at = NOW();
END;

• Row-level triggers only - no statement-level trigger support.
• No INSTEAD OF triggers on views.
• No DDL triggers.
• Multiple triggers per event/timing combination supported since MySQL 5.7.
• NEW and OLD row references available in trigger body.

CREATE TRIGGER trg_set_updated_at
BEFORE UPDATE ON orders
FOR EACH ROW
BEGIN
  SET NEW.updated_at = NOW();
END;

• Same capabilities as MySQL: BEFORE/AFTER row-level DML triggers.
• No statement-level or INSTEAD OF triggers.
• Sequence-aware triggers available in newer versions.
• Oracle-compatible trigger syntax available in sql_mode=ORACLE.

-- Row-level DML trigger
CREATE OR REPLACE TRIGGER trg_set_updated_at
BEFORE UPDATE ON orders
FOR EACH ROW
BEGIN
  :NEW.updated_at := SYSDATE;
END;
/

-- DDL trigger
CREATE OR REPLACE TRIGGER trg_audit_ddl
AFTER CREATE ON SCHEMA
BEGIN
  INSERT INTO ddl_log(event, obj)
  VALUES(ORA_SYSEVENT, ORA_DICT_OBJ_NAME);
END;
/

• Supports BEFORE, AFTER, and INSTEAD OF timing for DML.
• DDL triggers fire on CREATE, ALTER, DROP events at schema or database level.
• Database event triggers fire on STARTUP, SHUTDOWN, LOGON, LOGOFF, SERVERERROR.
• COMPOUND TRIGGER combines multiple timing points (BEFORE STATEMENT, BEFORE EACH ROW, etc.) in one object.
• FOLLOWS / PRECEDES clause controls ordering of multiple triggers on the same event.

-- DML trigger (AFTER)
CREATE TRIGGER trg_set_updated_at
ON orders
AFTER UPDATE
AS BEGIN
  UPDATE orders
  SET updated_at = GETDATE()
  FROM orders o
  INNER JOIN inserted i ON o.id = i.id;
END;

-- DDL trigger
CREATE TRIGGER trg_audit_ddl
ON DATABASE
AFTER CREATE_TABLE
AS BEGIN
  INSERT INTO ddl_log(event_type, event_data)
  SELECT EVENTDATA().value('(/EVENT_INSTANCE/EventType)[1]','NVARCHAR(100)'),
         CONVERT(NVARCHAR(MAX), EVENTDATA());
END;

• No BEFORE triggers - only AFTER and INSTEAD OF.
• INSERTED and DELETED virtual tables contain the new and old row versions.
• DDL triggers fire at DATABASE or ALL SERVER scope.
• Logon triggers fire on user authentication events.
• EVENTDATA() function returns XML with DDL event details.

CREATE TRIGGER trg_set_updated_at
AFTER UPDATE ON orders
FOR EACH ROW
WHEN NEW.updated_at = OLD.updated_at
BEGIN
  UPDATE orders SET updated_at = datetime('now')
  WHERE id = NEW.id;
END;

• Supports BEFORE, AFTER, and INSTEAD OF timing.
• Row-level only - no statement-level triggers.
• INSTEAD OF triggers on views enable updatable views.
• WHEN clause provides conditional execution.
• OLD and NEW row references available.

• Redshift does not support triggers.
• Use Amazon EventBridge rules with Lambda for event-driven reactions to Redshift changes.
• Stored procedure polling can simulate some trigger-like behavior.
• Amazon Redshift Data API combined with scheduled queries provides an alternative.

CREATE TRIGGER trg_set_updated_at
BEFORE UPDATE ON orders
REFERENCING NEW AS n
FOR EACH ROW
BEGIN ATOMIC
  SET n.updated_at = CURRENT TIMESTAMP;
END;

• Supports BEFORE, AFTER, and INSTEAD OF timing.
• FOR EACH ROW and FOR EACH STATEMENT granularity.
• REFERENCING clause names the NEW and OLD row/table correlation names.
• Multiple trigger events (INSERT OR UPDATE OR DELETE) in one CREATE TRIGGER statement.
• Transition variables (NEW, OLD) and transition tables (NEW_TABLE, OLD_TABLE) available.

• Snowflake does not support triggers.
• Snowflake Streams capture change data (CDC) on tables as an alternative.
• Snowflake Tasks can be scheduled or triggered by stream lag to process stream data.
• The Streams + Tasks pattern provides event-driven processing similar to triggers.

• BigQuery does not support triggers.
• Use Pub/Sub notifications on BigQuery table changes to invoke Cloud Functions or Cloud Run.
• Eventarc can route BigQuery events (job completion, table changes) to Cloud Functions.
• Scheduled queries can poll for changes on a recurring basis.