I set up a first Hive table with GZIP-compressed files:

CREATE EXTERNAL TABLE table_gzip (
    col1,
    col2,
    col3
)
ROW FORMAT DELIMITED,
  FIELDS TERMINATED BY ','
  LINES TERMINATED BY '\n'
LOCATION
  's3://bucket/files_gzip/';

Then I set up another Hive table with ORC format:

CREATE EXTERNAL TABLE table_orc (
    col1,
    col2,
    col3
)
STORED AS ORC
LOCATION
   's3://bucket/files_orc/';
ALTER TABLE table_orc SET tblproperties ("orc.compress" ="SNAPPY");

And then I uncompressed and recompressed from GZIP to ORC using this query:

INSERT OVERWRITE TABLE table_gzip SELECT * FROM table_orc

Once this query finished, I had new ORC-compressed files in 's3://bucket/files_orc/'. So far so good.

However when I looked at the files, they went from 500 1.2GiB files to 500 1.6GiB files.

What did I do wrong? Why are my ORC-SNAPPY compressed files larger than the original files? Is GZIP a better compression method?

Thanks for your time.