Paimon catalog
Introduction
Apache Gravitino provides the ability to manage Apache Paimon metadata.
Requirements
Builds with Apache Paimon 0.8.0
.
Catalog
Catalog capabilities
-
Works as a catalog proxy, supporting
FilesystemCatalog
. -
Supports DDL operations for Paimon schemas and tables.
-
Doesn't support
JdbcCatalog
andHiveCatalog
catalog backend now. -
Doesn't support alterSchema.
Catalog properties
Property name | Description | Default value | Required | Since Version |
---|---|---|---|---|
catalog-backend | Catalog backend of Gravitino Paimon catalog. Only supports filesystem now. | (none) | Yes | 0.6.0-incubating |
uri | The URI configuration of the Paimon catalog. thrift://127.0.0.1:9083 or jdbc:postgresql://127.0.0.1:5432/db_name or jdbc:mysql://127.0.0.1:3306/metastore_db . It is optional for FilesystemCatalog . | (none) | required if the value of catalog-backend is not filesystem . | 0.6.0-incubating |
warehouse | Warehouse directory of catalog. file:///user/hive/warehouse-paimon/ for local fs, hdfs://namespace/hdfs/path for HDFS or s3://{bucket-name}/path/ for S3 | (none) | Yes | 0.6.0-incubating |
authentication.type | The type of authentication for Paimon catalog backend, currently Gravitino only supports Kerberos and simple . | simple | No | 0.6.0-incubating |
authentication.kerberos.principal | The principal of the Kerberos authentication. | (none) | required if the value of authentication.type is Kerberos. | 0.6.0-incubating |
authentication.kerberos.keytab-uri | The URI of The keytab for the Kerberos authentication. | (none) | required if the value of authentication.type is Kerberos. | 0.6.0-incubating |
authentication.kerberos.check-interval-sec | The check interval of Kerberos credential for Paimon catalog. | 60 | No | 0.6.0-incubating |
authentication.kerberos.keytab-fetch-timeout-sec | The fetch timeout of retrieving Kerberos keytab from authentication.kerberos.keytab-uri . | 60 | No | 0.6.0-incubating |
Any properties not defined by Gravitino with gravitino.bypass.
prefix will pass to Paimon catalog properties and HDFS configuration. For example, if specify gravitino.bypass.table.type
, table.type
will pass to Paimon catalog properties.
Catalog operations
Please refer to Manage Relational Metadata Using Gravitino for more details.
Schema
Schema capabilities
-
Supporting createSchema, dropSchema, loadSchema and listSchema.
-
Supporting cascade drop schema.
-
Doesn't support alterSchema.
Schema properties
- Doesn't support specify location and store any schema properties when createSchema for FilesystemCatalog.
- Doesn't return any schema properties when loadSchema for FilesystemCatalog.
- Doesn't support store schema comment for FilesystemCatalog.
Schema operations
Please refer to Manage Relational Metadata Using Gravitino for more details.
Table
Table capabilities
-
Supporting createTable, purgeTable, alterTable, loadTable and listTable.
-
Supporting Column default value through table properties, such as
fields.{columnName}.default-value
, not column expression. -
Doesn't support dropTable.
-
Doesn't support table distribution and sort orders.
Gravitino Paimon Catalog does not support dropTable, because the dropTable in Paimon will both remove the table metadata and the table location from the file system and skip the trash, we should use purgeTable instead in Gravitino.
Paimon does not support auto increment column.
Table changes
- RenameTable
- AddColumn
- DeleteColumn
- RenameColumn
- UpdateColumnComment
- UpdateColumnNullability
- UpdateColumnPosition
- UpdateColumnType
- UpdateComment
- SetProperty
- RemoveProperty
Table partitions
- Only supports Identity partitions, such as
day
,hour
, etc.
Please refer to Paimon DDL Create Table for more details.
Table sort orders
- Doesn't support table sort orders.
Table distributions
- Doesn't support table distributions.
Table indexes
- Only supports primary key Index.
We cannot specify more than one primary key Index, and a primary key Index can contain multiple fields as a joint primary key.
Paimon Table primary key constraint should not be same with partition fields, this will result in only one record in a partition.
Table column types
Gravitino Type | Apache Paimon Type |
---|---|
Struct | Row |
Map | Map |
List | Array |
Boolean | Boolean |
Byte | TinyInt |
Short | SmallInt |
Integer | Int |
Long | BigInt |
Float | Float |
Double | Double |
Decimal | Decimal |
String | VarChar(Integer.MAX_VALUE) |
VarChar | VarChar |
FixedChar | Char |
Date | Date |
Time | Time |
TimestampType withZone | LocalZonedTimestamp |
TimestampType withoutZone | Timestamp |
Binary | Binary |
Fixed | VarBinary |
Gravitino doesn't support Paimon MultisetType
type.
Table properties
You can pass Paimon table properties to Gravitino when creating a Paimon table.
The Gravitino server doesn't allow passing the following reserved fields.
Configuration item | Description |
---|---|
comment | The table comment. |
owner | The table owner. |
bucket-key | The table bucket-key. |
merge-engine | The table merge-engine. |
sequence.field | The table sequence.field. |
rowkind.field | The table rowkind.field. |
primary-key | The table primary-key. |
partition | The table partition. |
Table operations
Please refer to Manage Relational Metadata Using Gravitino for more details.
HDFS configuration
You can place core-site.xml
and hdfs-site.xml
in the catalogs/lakehouse-paimon/conf
directory to automatically load as the default HDFS configuration.
When reading and writing to HDFS, the Gravitino server can only operate as the specified Kerberos user and doesn't support proxying to other Kerberos users now.