Skip to content
BENIBIX Learn

Apache ORC

Also known as ORC, Optimized Row Columnar, ORC File

An open, column-based file format from the Hive world. It is close to Parquet.

Think of it like

It is like Parquet's cousin. It packs a box the same way, with a few small changes.

Example

A bank's old Hadoop cluster stores its tables as ORC. New tools on the same cluster can still read those files.

Why it matters

Many older data lakes are full of ORC files. You will meet it when you work on them.

Definition

Apache ORC, short for Optimized Row Columnar, is an open file format that stores data by column. It grew up inside the Hive project on Hadoop. A file is cut into large blocks called stripes, each with its own index. Like Parquet, it lets readers skip data they do not need. It is still common where Hive is in use.

Common mistake

Converting every ORC file to Parquet for no reason. Both work well, and the copy costs time and money.

Key takeaways

  1. 1
    By column. Columns are stored apart.
  2. 2
    Stripes. Big blocks, each with an index.
  3. 3
    Hive roots. Common in Hadoop setups.