Skip to main content

Iceberg Table Engine

This engine provides a read-only integration with existing Apache Iceberg tables in Amazon S3, Azure, HDFS and locally stored tables.

Create Table​

Note that the Iceberg table must already exist in the storage, this command does not take DDL parameters to create a new table.

CREATE TABLE iceberg_table_s3
    ENGINE = IcebergS3(url,  [, NOSIGN | access_key_id, secret_access_key, [session_token]], format, [,compression])

CREATE TABLE iceberg_table_azure
    ENGINE = IcebergAzure(connection_string|storage_account_url, container_name, blobpath, [account_name, account_key, format, compression])

CREATE TABLE iceberg_table_hdfs
    ENGINE = IcebergHDFS(path_to_table, [,format] [,compression_method])

CREATE TABLE iceberg_table_local
    ENGINE = IcebergLocal(path_to_table, [,format] [,compression_method])

Engine arguments

Description of the arguments coincides with description of arguments in engines S3, AzureBlobStorage, HDFS and File correspondingly. format stands for the format of data files in the Iceberg table.

Engine parameters can be specified using Named Collections

Example

CREATE TABLE iceberg_table ENGINE=IcebergS3('http://test.s3.amazonaws.com/clickhouse-bucket/test_table', 'test', 'test')

Using named collections:

<clickhouse>
    <named_collections>
        <iceberg_conf>
            <url>http://test.s3.amazonaws.com/clickhouse-bucket/</url>
            <access_key_id>test</access_key_id>
            <secret_access_key>test</secret_access_key>
        </iceberg_conf>
    </named_collections>
</clickhouse>
CREATE TABLE iceberg_table ENGINE=IcebergS3(iceberg_conf, filename = 'test_table')

Aliases

Table engine Iceberg is an alias to IcebergS3 now.

Data cache​

Iceberg table engine and table function support data caching same as S3, AzureBlobStorage, HDFS storages. See here.

See also​