Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Wrong (similar) field is returned

See original GitHub issue

Hadoop: 2.6 PrestoDB: Presto CLI 0.148 Connector: Hive (Hive 1.2)

I have a Parquet Hive table, with 20 fields:

fullname
browser_fullname
…

Running the query SELECT COUNT(*), fullname FROM sample2 GROUP BY fullname;

gives browser_fullname field instead of fullname.

Tested with Hive on Tez and Apache Drill, it’s ok:

Issue Analytics

State:
Created 7 years ago
Comments:5 (1 by maintainers)

Top GitHub Comments

5reactions

nezihyigitbasicommented, Mar 1, 2017

I suspect there is some sort of schema mismatch here, you should compare the Hive schema vs. the Parquet file level schema to confirm that.

BTW by default Presto uses index-based access when reading columns from Parquet files, and probably the Hive column index for the fullname column points to the browser_fullname column in the Parquet file. You can try setting hive.parquet.use-column-names=true to see whether it helps.

0reactions

ebuildycommented, Mar 21, 2017

Can you try with SET SESSION such as described https://prestodb.io/docs/current/sql/set-session.html ?