Apache Iceberg

1 min read

On this page

Background Information
Features
Limits
Syntax
Parameters in the WITH Clause
- Common Parameters
- Parameters Only for Result Tables
Data Type Mappings
Sample Code
- Sample Code for an Apache Iceberg Source Read Script
- Sample Code for an Apache Iceberg Sink Write Script

This topic describes how to use the Apache Iceberg connector.

Background Information

Apache Iceberg is an open table format for data lakes. You can use Apache Iceberg to quickly build your own data lake storage service on Hadoop Distributed File System (HDFS) or Amazon Simple Storage Service (S3). Then, you can use a computing engine of the open source big data ecosystem, such as Apache Flink, Apache Spark, Apache Hive, or Apache Presto, to analyze data in your data lake.

Item	Description
Item	Description
Table type	Source table and result table
Running mode	Streaming mode
Data format	N/A
Metric	N/A
Data update or deletion in a result table	Supported

Features

Apache Iceberg provides the following core capabilities:

Builds a low-cost lightweight data lake storage service based on HDFS or S3.
Provides comprehensive atomicity, consistency, isolation, durability (ACID) semantics.
Supports historical version backtracking.
Supports efficient data filtering.
Supports schema evolution.
Supports partition evolution.

Note

You can use the efficient fault tolerance and stream processing capabilities of Flink to import a large amount of behavioral data in logs into an Apache Iceberg data lake in real time. Then, you can use Flink or another analytics engine to extract the value of your data.

Limits

Only Ververica Cloud that uses VERA 1.0.3 and later supports the Apache Iceberg connector. The Apache Iceberg connector supports only the Apache Iceberg table format of version 1. For more information, see Iceberg Table Spec.

Syntax

SQL

1 CREATE TABLE iceberg_table (
2  id    BIGINT,
3  data  STRING
4) WITH (
5  'connector' = 'iceberg',
6  ...
7);

Parameters in the WITH Clause

Common Parameters

Parameter	Description	Data type	Required	Default value	Remarks
Parameter	Description	Data type	Required	Default value	Remarks
connector	The type of the source table.	STRING	Yes	No default value	Set the value to iceberg.
catalog-name	The name of the catalog.	STRING	Yes	No default value	Set the value to a custom name.
catalog-database	The name of the database.	STRING	Yes	default	Set the value to the name of the database that is created with Glue catalog.
io-impl	The name of the implementation class in the distributed file system.	STRING	Yes	No default value	Set the value to org.apache.iceberg.aws.s3.S3FileIO.
catalog-impl	The class name of the catalog.	STRING	Yes	No default value
warehouse	The S3 directory in which table data is stored.	STRING	Yes	No default value	N/A.

Parameters Only for Result Tables

Parameter	Description	Data type	Required	Default value	Remarks
Parameter	Description	Data type	Required	Default value	Remarks
write.operation	The write operation mode.	STRING	No	upsert	upsert: Data is updated. This is the default value.insert: Data is written to the table in append mode.bulk_insert: A specific amount of data is written at a time and existing data is not updated.
hive_sync.enable	Specifies whether to enable the synchronization of metadata to Hive.	BOOLEAN	No	false	Valid values: true: The synchronization of metadata to Hive is enabled.false: The synchronization of metadata to Hive is disabled. This is the default value.
hive_sync.mode	The Hive data synchronization mode.	STRING	No	hms	hms: If you use a Hive metastore catalog, retain the default value. This is the default value.jdbc: If the Java Database Connectivity (JDBC) catalog is used, set this value to jdbc.
hive_sync.db	The name of the Hive database to which data is synchronized.	STRING	No	Database name of the current table in the catalog	N/A.
hive_sync.table	The name of the Hive table to which data is synchronized.	STRING	No	Name of the current table	N/A.

Data Type Mappings

Data type of Apache Iceberg	Data type of Flink
Data type of Apache Iceberg	Data type of Flink
BOOLEAN	BOOLEAN
INT	INT
LONG	BIGINT
FLOAT	FLOAT
DOUBLE	DOUBLE
DECIMAL(P,S)	DECIMAL(P,S)
DATE	DATE
TIME	TIME Note: Apache Iceberg timestamps are accurate to the microsecond, and Flink timestamps are accurate to the millisecond. When you use Flink to read Apache Iceberg data, the time precision is aligned to milliseconds.
TIMESTAMP	TIMESTAMP
TIMESTAMPTZ	TIMESTAMP_LTZ
STRING	STRING
FIXED(L)	BYTES
BINARY	VARBINARY
STRUCT<...>	ROW
LIST<E>	LIST
MAP<K,V>	MAP

Sample Code

Create an S3 bucket.
Externally create an Iceberg catalog using the above S3 location.
Add a Glue policy to your IAM role (if using Glue as backend catalog).
Run the scripts.

Sample Code for an Apache Iceberg Source Read Script

SQL

1CREATE TEMPORARY TABLE iceberg_catalog_table (
2    id   INT,
3    data STRING
4) WITH (
5    'connector'='iceberg',
6    'catalog-name'='iceberg_db',
7    'catalog-database'='iceberg_db',
8    'catalog-table'='sample',
9    'catalog-impl'='org.apache.iceberg.aws.glue.GlueCatalog',
10    'io-impl'='org.apache.iceberg.aws.s3.S3FileIO',
11    'warehouse'='s3://iceberg-testing-us-west-1/catalog/'
12);
13
14CREATE TEMPORARY TABLE sink (
15  id   INT,
16  data STRING
17) WITH (
18  'connector' = 'blackhole'
19);
20
21INSERT INTO sink SELECT id, data
22FROM iceberg_catalog_table;

Sample Code for an Apache Iceberg Sink Write Script

SQL

1CREATE TEMPORARY TABLE datagen (
2    name VARCHAR,
3    age INT
4) WITH (
5    'connector' = 'datagen'
6);
7
8CREATE TEMPORARY TABLE iceberg_catalog_table (
9    id   INT,
10    data STRING
11) WITH (
12    'connector'='iceberg',
13    'catalog-name'='iceberg_db',
14    'catalog-database'='iceberg_db',
15    'catalog-table'='sample',
16    'catalog-impl'='org.apache.iceberg.aws.glue.GlueCatalog',
17    'io-impl'='org.apache.iceberg.aws.s3.S3FileIO',
18    'warehouse'='s3://iceberg-testing-us-west-1/catalog/',
19    's3.staging-dir'='s3://iceberg-testing-us-west-1/staging/'
20);
21
22INSERT INTO iceberg_catalog_table SELECT age, name
23FROM datagen;

Was this helpful?

Yes No