This patch set adds a couple of scripts that makes it easy to run
Trino in the dev environment.
At first one 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
One 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 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)
Testing
* no tests added
* later we might always run Trino as part of the minicluster and
add interop tests with it
Change-Id: I49818c7a95e23988b3fbc3d31b4c7fa738e0d952
Reviewed-on: http://gerrit.cloudera.org:8080/20444
Reviewed-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
Tested-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
cf819f967b