Oracle connector

The Oracle connector allows querying and creating tables in an external Oracle database. Connectors let Trino join data provided by different databases, like Oracle and Hive, or different Oracle database instances.

Requirements

To connect to Oracle, you need:

Oracle 12 or higher.
Network access from the Trino coordinator and workers to Oracle. Port 1521 is the default port.

Configuration

To configure the Oracle connector as the oracle catalog, create a file named oracle.properties in etc/catalog. Include the following connection properties in the file:

connector.name=oracle
# The correct syntax of the connection-url varies by Oracle version and
# configuration. The following example URL connects to an Oracle SID named
# "orcl".
connection-url=jdbc:oracle:thin:@example.net:1521:orcl
connection-user=root
connection-password=secret

The connection-url defines the connection information and parameters to pass to the JDBC driver. The Oracle connector uses the Oracle JDBC Thin driver, and the syntax of the URL may be different depending on your Oracle configuration. For example, the connection URL is different if you are connecting to an Oracle SID or an Oracle service name. See the Oracle Database JDBC driver documentation for more information.

The connection-user and connection-password are typically required and determine the user credentials for the connection, often a service user. You can use secrets to avoid actual values in the catalog properties files.

note

Oracle does not expose metadata comment via REMARKS column by default in JDBC driver. You can enable it using oracle.remarks-reporting.enabled config option. See Additional Oracle Performance Extensions for more details.

By default, the Oracle connector uses connection pooling for performance improvement. The below configuration shows the typical default values. To update them, change the properties in the catalog configuration file:

oracle.connection-pool.max-size=30
oracle.connection-pool.min-size=1
oracle.connection-pool.inactive-timeout=20m

To disable connection pooling, update properties to include the following:

oracle.connection-pool.enabled=false

Multiple Oracle servers

If you want to connect to multiple Oracle servers, configure another instance of the Oracle connector as a separate catalog.

To add another Oracle catalog, create a new properties file. For example, if you name the property file sales.properties, Trino creates a catalog named sales.

General configuration properties

The following table describes general catalog configuration properties for the connector:

Property name	Description	Default value
`case-insensitive-name-matching`	Support case insensitive schema and table names.	`false`
`case-insensitive-name-matching.cache-ttl`		`1m`
`case-insensitive-name-matching.config-file`	Path to a name mapping configuration file in JSON format that allows Trino to disambiguate between schemas and tables with similar names in different cases.	`null`
`case-insensitive-name-matching.refresh-period`	Frequency with which Trino checks the name matching configuration file for changes.	`0` (refresh disabled)
`metadata.cache-ttl`	Duration for which metadata, including table and column statistics, is cached.	`0` (caching disabled)
`metadata.cache-missing`	Cache the fact that metadata, including table and column statistics, is not available	`false`
`metadata.cache-maximum-size`	Maximum number of objects stored in the metadata cache	`10000`
`write.batch-size`	Maximum number of statements in a batched execution. Do not change this setting from the default. Non-default values may negatively impact performance.	`1000`

Procedures

system.flush_metadata_cache()
Flush JDBC metadata caches. For example, the following system call flushes the metadata caches for all schemas in the example catalog
```
USE example.myschema;
CALL system.flush_metadata_cache();
```

Case insensitive matching

When case-insensitive-name-matching is set to true, Trino is able to query non-lowercase schemas and tables by maintaining a mapping of the lowercase name to the actual name in the remote system. However, if two schemas and/or tables have names that differ only in case (such as "customers" and "Customers") then Trino fails to query them due to ambiguity.

In these cases, use the case-insensitive-name-matching.config-file catalog configuration property to specify a configuration file that maps these remote schemas/tables to their respective Trino schemas/tables:

{
  "schemas": [
    {
      "remoteSchema": "CaseSensitiveName",
      "mapping": "case_insensitive_1"
    },
    {
      "remoteSchema": "cASEsENSITIVEnAME",
      "mapping": "case_insensitive_2"
    }],
  "tables": [
    {
      "remoteSchema": "CaseSensitiveName",
      "remoteTable": "tablex",
      "mapping": "table_1"
    },
    {
      "remoteSchema": "CaseSensitiveName",
      "remoteTable": "TABLEX",
      "mapping": "table_2"
    }]
}

Queries against one of the tables or schemes defined in the mapping attributes are run against the corresponding remote entity. For example, a query against tables in the case_insensitive_1 schema is forwarded to the CaseSensitiveName schema and a query against case_insensitive_2 is forwarded to the cASEsENSITIVEnAME schema.

At the table mapping level, a query on case_insensitive_1.table_1 as configured above is forwarded to CaseSensitiveName.tablex, and a query on case_insensitive_1.table_2 is forwarded to CaseSensitiveName.TABLEX.

By default, when a change is made to the mapping configuration file, Trino must be restarted to load the changes. Optionally, you can set the case-insensitive-name-mapping.refresh-period to have Trino refresh the properties without requiring a restart:

case-insensitive-name-mapping.refresh-period=30s

Non-transactional INSERT

The connector supports adding rows using INSERT statements </sql/insert>. By default, data insertion is performed by writing data to a temporary table. You can skip this step to improve performance and write directly to the target table. Set the insert.non-transactional-insert.enabled catalog property or the corresponding non_transactional_insert catalog session property to true.

Note that with this property enabled, data can be corrupted in rare cases where exceptions occur during the insert operation. With transactions disabled, no rollback can be performed.

Querying Oracle

The Oracle connector provides a schema for every Oracle database.

Run SHOW SCHEMAS to see the available Oracle databases:

SHOW SCHEMAS FROM oracle;

If you used a different name for your catalog properties file, use that catalog name instead of oracle.

note

The Oracle user must have access to the table in order to access it from Trino. The user configuration, in the connection properties file, determines your privileges in these schemas.

Examples

If you have an Oracle database named web, run SHOW TABLES to see the tables it contains:

SHOW TABLES FROM oracle.web;

To see a list of the columns in the clicks table in the web database, run either of the following:

DESCRIBE oracle.web.clicks;
SHOW COLUMNS FROM oracle.web.clicks;

To access the clicks table in the web database, run the following:

SELECT * FROM oracle.web.clicks;

Type mapping

Both Oracle and Trino have types that are not supported by the Oracle connector. The following sections explain their type mapping.

Oracle to Trino type mapping

Trino supports selecting Oracle database types. This table shows the Oracle to Trino data type mapping:

Oracle to Trino type mapping
Oracle database type	Trino type	Notes
`NUMBER(p, s)`	`DECIMAL(p, s)`	See number
`NUMBER(p)`	`DECIMAL(p, 0)`	See number
`FLOAT[(p)]`	`DOUBLE`
`BINARY_FLOAT`	`REAL`
`BINARY_DOUBLE`	`DOUBLE`
`VARCHAR2(n CHAR)`	`VARCHAR(n)`
`VARCHAR2(n BYTE)`	`VARCHAR(n)`
`NVARCHAR2(n)`	`VARCHAR(n)`
`CHAR(n)`	`CHAR(n)`
`NCHAR(n)`	`CHAR(n)`
`CLOB`	`VARCHAR`
`NCLOB`	`VARCHAR`
`RAW(n)`	`VARBINARY`
`BLOB`	`VARBINARY`
`DATE`	`TIMESTAMP(0)`	See datetime
`TIMESTAMP(p)`	`TIMESTAMP`	See datetime
`TIMESTAMP(p) WITH TIME ZONE`	`TIMESTAMP WITH TIME ZONE`	See datetime

Trino to Oracle type mapping

Trino supports creating tables with the following types in an Oracle database. The table shows the mappings from Trino to Oracle data types:

note

For types not listed in the table below, Trino can't perform the CREATE TABLE <table> AS SELECT operations. When data is inserted into existing tables Oracle to Trino type mapping is used.

Trino to Oracle Type Mapping
Trino type	Oracle database type	Notes
`TINYINT`	`NUMBER(3)`
`SMALLINT`	`NUMBER(5)`
`INTEGER`	`NUMBER(10)`
`BIGINT`	`NUMBER(19)`
`DECIMAL(p, s)`	`NUMBER(p, s)`
`REAL`	`BINARY_FLOAT`
`DOUBLE`	`BINARY_DOUBLE`
`VARCHAR`	`NCLOB`
`VARCHAR(n)`	`VARCHAR2(n CHAR)` or `NCLOB`	See character
`CHAR(n)`	`CHAR(n CHAR)` or `NCLOB`	See character
`VARBINARY`	`BLOB`
`DATE`	`DATE`	See datetime
`TIMESTAMP`	`TIMESTAMP(3)`	See datetime
`TIMESTAMP WITH TIME ZONE`	`TIMESTAMP(3) WITH TIME ZONE`	See datetime

Mapping numeric types

An Oracle NUMBER(p, s) maps to Trino's DECIMAL(p, s) except in these conditions:

No precision is specified for the column (example: NUMBER or NUMBER(*)), unless oracle.number.default-scale is set.
Scale (s ) is greater than precision.
Precision (p ) is greater than 38.
Scale is negative and the difference between p and s is greater than 38, unless oracle.number.rounding-mode is set to a different value than UNNECESSARY.

If s is negative, NUMBER(p, s) maps to DECIMAL(p + s, 0).

For Oracle NUMBER (without precision and scale), you can change oracle.number.default-scale=s and map the column to DECIMAL(38, s).

Mapping datetime types

Selecting a timestamp with fractional second precision (p) greater than 3 truncates the fractional seconds to three digits instead of rounding it.

Oracle DATE type stores hours, minutes, and seconds, so it is mapped to Trino TIMESTAMP(0).

caution

Due to date and time differences in the libraries used by Trino and the Oracle JDBC driver, attempting to insert or select a datetime value earlier than 1582-10-15 results in an incorrect date inserted.

Mapping character types

Trino's VARCHAR(n) maps to VARCHAR2(n CHAR) if n is no greater than 4000. A larger or unbounded VARCHAR maps to NCLOB.

Trino's CHAR(n) maps to CHAR(n CHAR) if n is no greater than 2000. A larger CHAR maps to NCLOB.

Using CREATE TABLE AS to create an NCLOB column from a CHAR value removes the trailing spaces from the initial values for the column. Inserting CHAR values into existing NCLOB columns keeps the trailing spaces. For example:

CREATE TABLE vals AS SELECT CAST('A' as CHAR(2001)) col;
INSERT INTO vals (col) VALUES (CAST('BB' as CHAR(2001)));
SELECT LENGTH(col) FROM vals;

_col0
-------
 2001
    1
(2 rows)

Attempting to write a CHAR that doesn't fit in the column's actual size fails. This is also true for the equivalent VARCHAR types.

Type mapping configuration properties

The following properties can be used to configure how data types from the connected data source are mapped to Trino data types and how the metadata is cached in Trino.

Property name Description Default value

Property name	Description	Default value
`unsupported-type-handling`	Configure how unsupported column data types are handled: `IGNORE`, column is not accessible. `CONVERT_TO_VARCHAR`, column is converted to unbounded`VARCHAR`. The respective catalog session property is`unsupported_type_handling`.	`IGNORE`
`jdbc-types-mapped-to-varchar`	Allow forced mapping of comma separated lists of data types to convert to unbounded `VARCHAR`

unsupported-type-handling

Configure how unsupported column data types are handled:

IGNORE, column is not accessible.
CONVERT_TO_VARCHAR, column is converted to unboundedVARCHAR.

The respective catalog session property isunsupported_type_handling.

IGNORE

jdbc-types-mapped-to-varchar Allow forced mapping of comma separated lists of data types to convert to unbounded VARCHAR

Number to decimal configuration properties

Configuration property name	Session property name	Description	Default
`oracle.number.default-scale`	`number_default_scale`	Default Trino `DECIMAL` scale for Oracle`NUMBER` (without precision and scale) date type. When not set then such column is treated as not supported.	not set
`oracle.number.rounding-mode`	`number_rounding_mode`	Rounding mode for the Oracle `NUMBER` data type. This is useful when Oracle `NUMBER` data type specifies higher scale than is supported in Trino. Possible values are: `UNNECESSARY` - Rounding mode to assert that the requested operation has an exact result, hence no rounding is necessary. `CEILING` - Rounding mode to round towards positive infinity. `FLOOR` - Rounding mode to round towards negative infinity. `HALF_DOWN` - Rounding mode to round towards`nearest neighbor` unless both neighbors are equidistant, in which case rounding down is used. `HALF_EVEN` - Rounding mode to round towards the`nearest neighbor` unless both neighbors are equidistant, in which case rounding towards the even neighbor is performed. `HALF_UP` - Rounding mode to round towards`nearest neighbor` unless both neighbors are equidistant, in which case rounding up is used `UP` - Rounding mode to round towards zero. `DOWN` - Rounding mode to round towards zero.	`UNNECESSARY`

SQL support

The connector provides read access and write access to data and metadata in Oracle. In addition to the globally available and read operation statements, the connector supports the following statements:

INSERT
DELETE
TRUNCATE
CREATE TABLE
CREATE TABLE
DROP TABLE
ALTER TABLE
COMMENT

SQL DELETE

If a WHERE clause is specified, the DELETE operation only works if the predicate in the clause can be fully pushed down to the data source.

ALTER TABLE

The connector does not support renaming tables across multiple schemas. For example, the following statement is supported:

ALTER TABLE catalog.schema_one.table_one RENAME TO catalog.schema_one.table_two

The following statement attempts to rename a table across schemas, and therefore is not supported:

ALTER TABLE catalog.schema_one.table_one RENAME TO catalog.schema_two.table_two

Performance

The connector includes a number of performance improvements, detailed in the following sections.

Synonyms

Based on performance reasons, Trino disables support for Oracle SYNONYM. To include SYNONYM, add the following configuration property:

oracle.synonyms.enabled=true

Pushdown

The connector supports pushdown for a number of operations:

Pushdown
Pushdown
Pushdown

In addition, the connector supports Pushdown for the following functions:

avg()
count(), also count(distinct x)
max()
min()
sum()

Pushdown is only supported for DOUBLE type columns with the following functions:

stddev() and stddev_samp()
stddev_pop()
var_pop()
variance() and var_samp()

Pushdown is only supported for REAL or DOUBLE type column with the following functions:

covar_samp()
covar_pop()

Join pushdown

The join-pushdown.enabled catalog configuration property or join_pushdown_enabled catalog session property <session-properties-definition> control whether the connector pushes down join operations. The property defaults to false, and enabling join pushdowns may negatively impact performance for some queries.

Predicate pushdown support

The connector does not support pushdown of any predicates on columns that use the CLOB, NCLOB, BLOB, or RAW(n) Oracle database types, or Trino data types that map to these Oracle database types.

In the following example, the predicate is not pushed down for either query since name is a column of type VARCHAR, which maps to NCLOB in Oracle:

SHOW CREATE TABLE nation;

--             Create Table
----------------------------------------
-- CREATE TABLE oracle.trino_test.nation (
--    name varchar
-- )
-- (1 row)

SELECT * FROM nation WHERE name > 'CANADA';
SELECT * FROM nation WHERE name = 'CANADA';

In the following example, the predicate is pushed down for both queries since name is a column of type VARCHAR(25), which maps to VARCHAR2(25) in Oracle:

SHOW CREATE TABLE nation;

--             Create Table
----------------------------------------
-- CREATE TABLE oracle.trino_test.nation (
--    name varchar(25)
-- )
-- (1 row)

SELECT * FROM nation WHERE name > 'CANADA';
SELECT * FROM nation WHERE name = 'CANADA';

Oracle connector

Requirements​

Configuration​

Multiple Oracle servers​

General configuration properties​

Procedures​

Case insensitive matching​

Non-transactional INSERT​

Querying Oracle​

Examples​

Type mapping​

Oracle to Trino type mapping​

Trino to Oracle type mapping​

Mapping numeric types​

Mapping datetime types​

Mapping character types​

Type mapping configuration properties​

Number to decimal configuration properties​

SQL support​

SQL DELETE​

ALTER TABLE​

Performance​

Synonyms​

Pushdown​

Join pushdown​

Predicate pushdown support​

Requirements

Configuration

Multiple Oracle servers

General configuration properties

Procedures

Case insensitive matching

Non-transactional INSERT

Querying Oracle

Examples

Type mapping

Oracle to Trino type mapping

Trino to Oracle type mapping

Mapping numeric types

Mapping datetime types

Mapping character types

Type mapping configuration properties

Number to decimal configuration properties

SQL support

SQL DELETE

ALTER TABLE

Performance

Synonyms

Pushdown

Join pushdown

Predicate pushdown support