Databricks compatibility

This page compares Query Engine with Databricks for Unity Catalog managed Delta tables. Min. DBR is the first Databricks Runtime version with the feature. For table features, read means Query Engine can query a table that has the feature enabled, and write means it can write to one. For function-by-function support, see Databricks function compatibility.

Yes Supported Partial Supported with limits No Not supported yet

Summary

Area	Yes	Partial	No
Table features	23	4	11
Data types	17	1	7
DDL	0	0	42
DML	15	1	15
Metadata queries and time travel	4	1	5
Table maintenance	1	0	5
UDFs	0	0	6
Stored procedures and SQL scripting	0	0	6
Object-level access control	4	0	10
Attribute-based access control (ABAC)	3	0	9

Table features

A read-only engine is blocked by an unsupported Reader feature. Writer features only matter for DML.

Feature	Delta protocol	Min. DBR	Databricks	e6data read	e6data write
appendOnly	Writer	—	Yes	Yes	Yes
CHECK constraints	Writer	All	Yes	Yes	No
Generated columns	Writer	All	Yes	Yes	No
Identity columns	Writer	All	Yes	Yes	Partial
allowColumnDefaults	Writer	—	Yes	Yes	Yes
Change data feed	Writer	All	Yes	Yes	Yes
Column mapping (name / id)	Reader + Writer	All	Yes	Yes	Yes
Deletion vectors	Reader + Writer	All	Yes	Yes	No
TimestampNTZ	Reader + Writer	13.3	Yes	Yes	No
Liquid clustering	Writer	13.3	Yes	Yes	No
UniForm (Iceberg reads)	Writer	13.3	Yes	Yes	Yes
Row tracking	Writer	14.3	Yes	Yes	Yes
Domain metadata	Writer	—	Yes	Yes	Yes
V2 checkpoints	Reader + Writer	—	Yes	Yes	Yes
Type widening	Reader + Writer	15.4	Yes	Partial	No
Variant	Reader + Writer	15.4	Yes	Partial	No
Collations	—	16.1	Yes	Yes	No
Protected checkpoints	Writer	16.3	Yes	Partial	No
Catalog commits	Reader + Writer	16.4	Yes	No	No

Data types

Numeric

Data type	Min. DBR	Databricks	e6data
TINYINT	—	Yes	Yes
SMALLINT	—	Yes	Yes
INT	—	Yes	Yes
BIGINT	—	Yes	Yes
DECIMAL(p,s)	—	Yes	Yes
FLOAT	—	Yes	Yes
DOUBLE	—	Yes	Yes

String and binary

Data type	Min. DBR	Databricks	e6data
STRING	—	Yes	Yes
STRING COLLATE (collated string)	16.1	Yes	No
BINARY	—	Yes	No

Boolean and null

Data type	Min. DBR	Databricks	e6data
BOOLEAN	—	Yes	Yes
VOID	18.2	Yes	Yes

Date and time

Data type	Min. DBR	Databricks	e6data
DATE	—	Yes	Yes
TIME	—	Yes	Yes
TIMESTAMP	—	Yes	Yes
TIMESTAMP_NTZ	13.3	Yes	Yes

Interval

Data type	Min. DBR	Databricks	e6data
INTERVAL YEAR TO MONTH	—	Yes	No
INTERVAL DAY TO SECOND	—	Yes	No

Geospatial

Data type	Min. DBR	Databricks	e6data
GEOMETRY(srid)	17.1	Yes	No
GEOGRAPHY(srid)	17.1	Partial	No

Complex and semi-structured

Data type	Min. DBR	Databricks	e6data
ARRAY<elementType>	—	Yes	Yes
MAP<keyType, valueType>	—	Yes	Yes
STRUCT<fields>	—	Yes	Yes
VARIANT	15.4	Yes	Partial
OBJECT	—	Yes	No

DDL

Catalog and schema

Capability	Min. DBR	Databricks	e6data
CREATE / ALTER / DROP CATALOG	—	Yes	No
CREATE / ALTER / DROP SCHEMA	—	Yes	No
USE CATALOG / USE SCHEMA	—	Yes	No

CREATE TABLE

Capability	Min. DBR	Databricks	e6data
CREATE TABLE (managed)	—	Yes	No
CREATE TABLE with LOCATION (external)	—	Yes	No
CREATE TABLE IF NOT EXISTS	—	Yes	No
CREATE OR REPLACE TABLE	—	Yes	No
CREATE TABLE AS SELECT (CTAS)	—	Yes	No
CREATE TABLE LIKE	—	Yes	No
PARTITIONED BY	—	Yes	No
CLUSTER BY (liquid)	—	Yes	No
NOT NULL	—	Yes	No
DEFAULT column values	—	Yes	No
GENERATED ALWAYS AS (expr)	—	Yes	No
GENERATED AS IDENTITY	—	Yes	No
PRIMARY KEY / FOREIGN KEY	—	Yes	No
TBLPROPERTIES	—	Yes	No
Table and column COMMENT	—	Yes	No
CREATE TEMP TABLE	—	Yes	No
SHALLOW CLONE / DEEP CLONE	—	Yes	No

ALTER TABLE

Capability	Min. DBR	Databricks	e6data
RENAME TO	—	Yes	No
ADD COLUMNS (incl. nested)	—	Yes	No
DROP COLUMN	—	Yes	No
RENAME COLUMN	—	Yes	No
ALTER COLUMN TYPE (widening)	—	Yes	No
Reorder columns (FIRST / AFTER)	—	Yes	No
SET / DROP NOT NULL	—	Yes	No
SET / DROP DEFAULT	—	Yes	No
ADD / DROP CONSTRAINT (CHECK)	—	Yes	No
SET / UNSET TBLPROPERTIES	—	Yes	No
CLUSTER BY change / NONE	—	Yes	No
DROP FEATURE (protocol downgrade)	—	Yes	No
SET ROW FILTER / SET MASK	—	Yes	No
SET TAGS	—	Yes	No
OWNER TO	—	Yes	No

Views and table lifecycle

Capability	Min. DBR	Databricks	e6data
CREATE / ALTER / DROP VIEW	—	Yes	No
CREATE TEMP VIEW	—	Yes	No
Materialized views	—	Yes	No
DROP TABLE	—	Yes	No
UNDROP TABLE	—	Yes	No
TRUNCATE TABLE	—	Yes	No
COMMENT ON	—	Yes	No

DML

INSERT

Capability	Min. DBR	Databricks	e6data
INSERT INTO … VALUES	—	Yes	Yes
INSERT INTO … SELECT	—	Yes	Yes
INSERT with column list	—	Yes	Yes
INSERT … BY NAME	—	Yes	No
DEFAULT keyword in VALUES	—	Yes	No
INSERT OVERWRITE (full table)	—	Yes	No
INSERT OVERWRITE PARTITION (static / dynamic)	—	Yes	No
INSERT … REPLACE WHERE	12.2 LTS	Yes	No
INSERT … REPLACE USING	16.3	Yes	No
INSERT … REPLACE ON	17.1	Yes	No
INSERT WITH SCHEMA EVOLUTION	—	Yes	No
COPY INTO	—	Yes	No

UPDATE

Capability	Min. DBR	Databricks	e6data
UPDATE … WHERE	—	Yes	Yes
↳ CoW	—	Yes	Yes
↳ MoR (deletion vectors)	—	Yes	No
UPDATE nested struct fields	—	Yes	Partial

DELETE

Capability	Min. DBR	Databricks	e6data
DELETE … WHERE	—	Yes	Yes
↳ CoW	—	Yes	Yes
↳ MoR (deletion vectors)	—	Yes	No
DELETE with subquery in WHERE	—	Yes	Yes

MERGE

Capability	Min. DBR	Databricks	e6data
WHEN MATCHED THEN UPDATE	—	Yes	Yes
↳ CoW	—	Yes	Yes
↳ MoR (deletion vectors)	—	Yes	No
WHEN MATCHED THEN DELETE	—	Yes	Yes
↳ CoW	—	Yes	Yes
↳ MoR (deletion vectors)	—	Yes	No
WHEN NOT MATCHED THEN INSERT	—	Yes	Yes
WHEN NOT MATCHED BY SOURCE (UPDATE / DELETE)	—	Yes	Yes
Multiple conditional WHEN clauses	—	Yes	Yes
UPDATE SET * / INSERT *	—	Yes	No
MERGE WITH SCHEMA EVOLUTION	—	Yes	No

Metadata queries and time travel

Capability	Min. DBR	Databricks	e6data
DESCRIBE HISTORY	—	Yes	Yes
DESCRIBE DETAIL	—	Yes	Yes
DESCRIBE TABLE EXTENDED	—	Yes	No
SHOW TBLPROPERTIES	—	Yes	No
SHOW CREATE TABLE	—	Yes	No
SHOW CATALOGS / SCHEMAS / TABLES / COLUMNS	—	Yes	Partial
information_schema	—	Yes	No
Time travel: VERSION AS OF	—	Yes	Yes
Time travel: TIMESTAMP AS OF	—	Yes	Yes
table_changes() (CDF read)	—	Yes	No

Table maintenance

Capability	Min. DBR	Databricks	e6data
OPTIMIZE	—	Yes	No
OPTIMIZE … ZORDER BY	—	Yes	No
VACUUM	—	Yes	No
REORG TABLE … APPLY (PURGE)	—	Yes	No
RESTORE TABLE (version / timestamp)	—	Yes	No
ANALYZE TABLE COMPUTE STATISTICS	—	Yes	Yes

UDFs

Capability	Min. DBR	Databricks	e6data
SQL scalar UDF (persistent, UC)	—	Yes	No
SQL table-valued function	—	Yes	No
Python scalar UDF (UC)	—	Yes	No
Batch / pandas Python UDF (UC)	—	Yes	No
Python UDTF	—	Yes	No
Scala / Java UDF, UDAF	—	Yes	No

Stored procedures and SQL scripting

Capability	Min. DBR	Databricks	e6data
CREATE / CALL / DROP PROCEDURE	17.0	Yes	No
DESCRIBE / SHOW PROCEDURES	17.0	Yes	No
IN / OUT / INOUT parameters, defaults	17.0	Yes	No
SQL SECURITY INVOKER / DEFINER	17.0	Yes	No
SQL scripting (control flow, handlers)	—	Yes	No
Nested procedure calls	—	Yes	No

Object-level access control

Capability	Min. DBR	Databricks	e6data native	Inherited from Unity Catalog
GRANT / REVOKE on securables	—	Yes	No	Yes
Privilege inheritance (catalog → schema → table)	—	Yes	No	Yes
Dynamic views	—	Yes	No	No
Table-level row filters	—	Yes	No	Yes
Table-level column masks	—	Yes	No	Yes
Audit logs	—	Yes	No	No
Lineage	—	Yes	No	No

Attribute-based access control (ABAC)

Native = access rules defined and enforced inside e6data. Inherited from UC = rules defined in Unity Catalog and enforced by e6data when it queries a UC table.

Capability	Min. DBR	Databricks	e6data native	Inherited from Unity Catalog
Governed tags	—	Yes	No	Yes
Tag-based row filter policies	16.4	Yes	No	Yes
Tag-based column mask policies	16.4	Yes	No	No
Policies scoped at catalog / schema	—	Yes	No	Yes
Auto-apply to newly tagged tables	—	Yes	No	No
Automated data classification	—	Yes	No	No

Query Engine

Get started

Connect data

Run clusters

Query

Connect your tools

Secure and govern

Troubleshoot

Reference

Concepts

Databricks compatibility

Summary

Table features

Data types

Numeric

String and binary

Boolean and null

Date and time

Interval

Geospatial

Complex and semi-structured

DDL

Catalog and schema

CREATE TABLE

ALTER TABLE

Views and table lifecycle

DML

INSERT

UPDATE

DELETE

MERGE

Metadata queries and time travel

Table maintenance

UDFs

Stored procedures and SQL scripting

Object-level access control

Attribute-based access control (ABAC)

​Summary

​Table features

​Data types

​Numeric

​String and binary

​Boolean and null

​Date and time

​Interval

​Geospatial

​Complex and semi-structured

​DDL

​Catalog and schema

​CREATE TABLE

​ALTER TABLE

​Views and table lifecycle

​DML

​INSERT

​UPDATE

​DELETE

​MERGE

​Metadata queries and time travel

​Table maintenance

​UDFs

​Stored procedures and SQL scripting

​Object-level access control

​Attribute-based access control (ABAC)

Summary

Table features

Data types

Numeric

String and binary

Boolean and null

Date and time

Interval

Geospatial

Complex and semi-structured

DDL

Catalog and schema

CREATE TABLE

ALTER TABLE

Views and table lifecycle

DML

INSERT

UPDATE

DELETE

MERGE

Metadata queries and time travel

Table maintenance

UDFs

Stored procedures and SQL scripting

Object-level access control

Attribute-based access control (ABAC)