# Data Skew

> When one part of a split job gets far more rows than the rest, so all wait.

**Field:** Data Engineering  
**Category:** Processing  
**Also known as:** Partition Skew, Skewed Data, Key Skew

## Think of it like

It is like five friends sharing chores, but one gets all the dishes. Everyone waits for that one friend.

## Definition

Data skew is when one part of a split job gets far more data than the others. Big jobs split rows into parts by a key, such as country. If most rows share one key, one machine gets most of the work. The rest finish fast and then sit waiting. The whole job is only as quick as its slowest part.

## Example

A job groups sales by country on 50 machines. 80% of the sales come from one country, so one machine does most of the work while 49 wait.

## Why it matters

It is a common reason a big job runs for hours instead of minutes. Adding machines does not help.

## Common mistake

Adding more machines to fix it. The busy part stays the same size, so the job is no faster.

## Three principles

- **Uneven parts.** One key holds most of the rows.
- **Slowest part wins.** The job ends when the last part ends.
- **Split the busy key.** Add a random suffix to spread it out.

## Related terms

- [Data Shuffle](https://learn.benibix.com/term/data/data-shuffle)
- [Apache Spark](https://learn.benibix.com/term/data/apache-spark)
- [Data Partitioning](https://learn.benibix.com/term/data/data-partitioning)
- [Massively Parallel Processing (MPP)](https://learn.benibix.com/term/data/massively-parallel-processing)
- [MapReduce](https://learn.benibix.com/term/data/mapreduce)

## Told apart from

- [Data Skew vs Data Shuffle](https://learn.benibix.com/compare/data-skew-vs-data-shuffle): People think data skew and a data shuffle are the same slowdown, because both show up when a big job on many machines is slow. A shuffle is the step where rows move between machines, so rows with the same key meet. Skew is when one key has far more rows than the rest, so one machine gets most of them. A shuffle is often needed. Skew is a problem you can fix.

![Four workers split rows by country. Worker 1 has France with 5 percent, worker 3 Germany with 6 percent and worker 4 the UK and others with 9 percent, and all three are done. Worker 2 has a long highlighted bar for the US with 80 percent, and a dashed line at its end marks where the job ends. A note says three workers sit idle, and splitting the busy key evens out the work.](https://learn.benibix.com/diagrams/dat-data-skew.svg)

---

Source: https://learn.benibix.com/term/data/data-skew
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-19
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
