CREATE ICEBERG TABLE

Syntax

CREATE ICEBERG TABLE <table_identifier>
([ <column_name> <column_type> [, ...]])
    { PARTITION | CLUSTER } BY <column_name> [, ...]
    PRIMARY KEY <column_name> [, ...]
    -- Specify sort order (default is ASC)
    ORDER BY <column_name> [ASC | DESC] [, ...] -
    [{ table_options }];

Table identifier

Table identifiers are provided in the following format:

<catalog_name>.<schema_name>.<table_name>

Note that only metastore connection types are accepted for the catalog name.

Valid table names match the following identifier format:

identifier = "([^"]|"")*"|[A-Za-z_][A-Za-z0-9_]*;

Column types

COLUMN_TYPE = { DATE 
              | TIMESTAMP 
              | BIGINT 
              | DOUBLE 
              | STRING 
              | BOOLEAN }

Partition/Cluster clause

{ PARTITION | CLUSTER } BY <column_name> [, ...]

Upsolver enhances Iceberg's core functionality by supporting two data organization strategies: Partitioning and Clustering. Partitioning is a native feature of Iceberg, allowing efficient data organization to significantly improve query performance by optimizing how data is stored and accessed. Clustering, an additional feature provided by Upsolver, further optimizes performance for high-cardinality datasets by organizing related data rows within larger files.

Note: You can choose either PARTITION BY or CLUSTER BY for your table, allowing you to select the strategy that best fits your data's structure and usage patterns.

General

Partitions\Clusters can be defined at table creation or later with ALTER command.
Supported column types include string, bigint, Boolean, and date.
For staging tables ingesting external data, partition columns must match incoming field names exactly and be at the root level.
If this is a target table that you intend on writing an aggregated output to, the partition column should be one of the aggregation key columns and not an aggregated column.

Partitioning

PARTITION  BY <column_name> [, ...]

Each unique value within the specified columns creates a separate partition in the underlying storage, allowing queries to read only the files associated with a specific partition. This reduces the need to scan the entire dataset.

For more information, see Working with partition indexes.

Clustering

CLUSTER BY <column_name> [, ...]

Adaptive clustering merges small files into a larger, non-partitioned file, reducing the overall file count while maintaining an organized structure. Clustering is ideal for high-cardinality datasets or when multiple partitions are being written to in parallel. Periodically, merged files may be moved into appropriate partitions if they represent a significant portion of data.

For details on usage and benefits, refer to our Adaptive Clustering article.

Primary key clause

PRIMARY KEY <column_name> [, ...]

A table's primary key column(s) contains the values that uniquely identify each row.

When using an INSERT job to write to a table, values are inserted or updated based on the primary keys. This means that when writing to a table with no primary keys defined, the job only ever appends the new data that arrives; no existing values are updated.

In order to use a table in a MERGE job, at least one primary key column should be defined.

Additionally, if this is a staging table that you intend on using to ingest external data with COPY FROM, you should not define any primary key columns.

Order by clause

The data within a partition can be sorted by column(s) in order to gain performance. To do this, include the ORDER KEY clause, followed by the name of the column(s).

-- Specify sort order (default is ASC)
ORDER BY <column_name> [ ASC | DESC ] [, ...]

For more information, please refer to the Apache Iceberg documentation.

The ORDER BY clause allows Upsolver to sort data within partitions by columns to gain performance.

The sort order is defined by a list of sort fields. The order of the sort fields within the list defines the order in which the sort is applied to the data. Each sort field has a sort direction, that can only be either ASC or DESC.

Table options

[ COMMENT = '<comment>' ]
[ STORAGE_CONNECTION = <connection_identifier>
  STORAGE_LOCATION = '<storage_location>' ]
[ COMPUTE_CLUSTER = <cluster_identifier> ]
[ DISABLE_COMPACTION = { TRUE | FALSE } ]
[ RETENTION_COLUMN = <column_name> ]
[ RETENTION_DURATION = <integer> DAYS ]
[ iceberg.<some.iceberg.property> = '<property_value>' ]

Jump to

`COMMENT` — editable

Type: text

(Optional) A description or comment regarding this table.

`STORAGE_CONNECTION`

Type: identifier

Default: Default storage connection configured for the metastore connection this table is created under

(Optional) The storage connection associated with the STORAGE_LOCATION for the table's underlying files.

Only a storage type connection can be used here (e.g. S3, Blob storage, GCS, Oracle object storage), and it should match the catalog's metastore. For example, if Glue is used as the metastore, only S3 is allowed as a storage connection.

When set, STORAGE_LOCATION must be configured as well to provide a path to store the data.

`STORAGE_LOCATION`

Type: text

Default: Default storage location configured for the metastore connection this table is created under

(Optional) The storage location for the table's underlying files.

For S3, it should be provided in the format s3://bucket_name/path_to_data.

This option is required when STORAGE_CONNECTION is set.

When set, STORAGE_CONNECTION must be configured as well to provide a connection with access to write to the specified storage location.

`COMPUTE_CLUSTER` — editable

Type: identifier

Default: The sole cluster in your environment

(Optional) The compute cluster that processes the table.

This option can only be omitted when there is just one cluster in your environment. You must specify which one to use when you have more than one compute cluster.

`DISABLE_COMPACTION` — editable

Type: Boolean

Default: false

(Optional) When true, disables the compaction process.

`RETENTION_COLUMN`

Type: identifier

(Optional) Data will be deleted from the table based on values in this column.

The column type must be DATE, TIMESTAMP, TIMESTAMPTZ, LONG, or INT.

For performance and cost reasons, it is preferable that the RETENTION_COLUMN is part of the partition columns defined for the table. However, in some use cases, you may want to select a retention column that is not part of your partition columns. In such cases, deletion occurs at the file level. Therefore, data will be deleted only once all rows in the file have passed the RETENTION_DURATION.

If you select a column of type LONG or INT, Upsolver will automatically detect the unit (seconds, milliseconds, microseconds, minutes, days) in which the data is stored.

This option is stored in the Iceberg table properties under the key upsolver.retention.column.

`RETENTION_DURATION` — editable

Value: <integer> DAYS

(Optional) Data must be at least this many days old to be deleted. Number of days can range between 1 and 9999.

This option functions similarly to lifecycle policies on common blob storage services, such as Amazon S3, and is designed to save storage costs. While it does not delete data immediately upon surpassing the defined threshold, it ensures that data will be deleted at some point in the future. Once data passes the retention period, it will be scheduled for deletion, and although the exact time may vary, Upsolver aims to complete this process within a reasonable timeframe.

This option is stored in the Iceberg table properties under the key 'upsolver.retention.duration'.

`ICEBERG_TABLE_PROPERTIES`

You can configure any Iceberg table property as documented in the Apache Iceberg table properties documentation.

iceberg.<some.iceberg.property> = '<property_value>'

Note:

The prefix iceberg. should be added before the name of the property.
Properties which include the character "-" should be written within double quotes. For example: iceberg.read.split."planning-lookback" = '10'
Iceberg table properties are strings, so the <property_value> should be enclosed in apostrophes.

Examples

Minimum example

CREATE ICEBERG TABLE default_glue_catalog.my_database.my_iceberg_table();

Partitioned table example

CREATE ICEBERG TABLE default_glue_catalog.my_database.my_iceberg_table()
    PARTITIONED BY $event_date;

Basic example

CREATE ICEBERG TABLE default_glue_catalog.my_database.orders_data
(
    order_id string,
    order_date date,
    customer_email string,
    net_total bigint, 
    num_items bigint,
    customer_address string,
    customer_city string,
    customer_state string,
    customer_zipcode bigint
)
    PARTITIONED BY order_date
    RETENTION_COLUMN = order_date
    RETENTION_DURATION = 7 DAYS
    PRIMARY KEY order_id
    ORDER BY net_total DESC, order_date ASC
    STORAGE_CONNECTION = s3_connection
    STORAGE_LOCATION = 's3://bucket/storage_location'
    COMPUTE_CLUSTER = "my cluster"
    DISABLE_COMPACTION = false
    iceberg.read.split."planning-lookback" = '10'
    COMMENT = 'Orders table';

Last updated 24 days ago

Syntax

Jump to

Table identifier

Column types

Partition/Cluster clause

Primary key clause

Order by clause

Table options

COMMENT — editable

STORAGE_CONNECTION

STORAGE_LOCATION

COMPUTE_CLUSTER — editable

DISABLE_COMPACTION — editable

RETENTION_COLUMN

RETENTION_DURATION — editable

ICEBERG_TABLE_PROPERTIES