Setting up Trino in the development environment

Impala has a couple of scripts that make it easy to run Trino in the dev environment.

First we need to build our custom Trino docker image via: testdata/bin/build-trino-docker-image.sh

Then we can run the Trino container via: testdata/bin/run-trino.sh

We can connect to the Trino CLI by: testdata/bin/trino-cli.sh

Trino will be configured to access our HMS and HDFS. We add the ‘hive’ and ‘iceberg’ catalogs for Trino. Legacy Hive tables can be accessed via the ‘hive’ catalog, while Iceberg tables (only the ones that reside in HiveCatalog) can be accessed via the ‘iceberg’ catalog. E.g.:

trino> use iceberg.functional_parquet;
trino:functional_parquet> select count(*) from iceberg_mixed_file_format;
 _col0
-------
     3
(1 row)