Dear Apache Iceberg team,
The Apache Impala team is pleased to announce official Apache Iceberg
support with the release of Impala 4.1.0. Impala can create/read/write V1
tables and the performance matches the native Parquet tables in many cases.
Impala has SQL syntax support for V1 features like partition transforms,
partition and schema evolution, time travel, and so on. The syntax is
mostly aligned with Hive’s syntax. E.g.:
CREATE TABLE ice_t (s STRING, i INT, j INT)
PARTITIONED BY SPEC (TRUNCATE(1, s))
STORED AS ICEBERG;
INSERT INTO ice_t VALUES ('iceberg', 1, 11);
ALTER TABLE ice_t SET PARTITION SPEC (void(s), BUCKET(7, s), j);
ALTER TABLE ice_t CHANGE COLUMN i big_i BIGINT;
INSERT INTO ice_t VALUES ('impala', 5, 5);
DESCRIBE HISTORY ice_t;
SELECT * FROM ice_t;
SELECT * FROM ice_t FOR SYSTEM_TIME AS OF '2022-06-17 11:18:19';
SELECT * FROM ice_t FOR SYSTEM_TIME AS OF now() - interval 1 days;
SELECT * FROM ice_t FOR SYSTEM_VERSION AS OF 860514146484624738;
The release is available at: https://impala.apache.org/downloads.html
Impala Iceberg documentation is available at:
https://impala.apache.org/docs/build/html/topics/impala_iceberg.html
We would also like to add Impala to the list of engines that supports
Iceberg, the following doc PRs have been created:
https://github.com/apache/iceberg/pull/5076
https://github.com/apache/iceberg-docs/pull/94
Please try Impala and share your feedback on either d...@impala.apache.org
or <d...@impala.apache.org>u...@impala.apache.org. Also, please feel free to
open Jiras about feature requests / bug reports at
https://issues.apache.org/jira/projects/IMPALA/ (with label
‘impala-iceberg’).
Thanks,
The Apache Impala team
