---
metadata:
  - name: generator
    content: Diplodoc Platform v5.52.0
alternate:
  - https://catboost.ai/docs/en/concepts/spark-cluster-configuration.md
  - href: en/concepts/spark-cluster-configuration.md
    type: text/markdown
    title: Markdown version
  - href: ../llms.txt
    type: text/markdown
    title: llms.txt
---
> **Documentation Index:** Fetch the complete configuration index at https://catboost.ai/docs/en/llms.txt

# Spark cluster configuration


CatBoost for Apache Spark requires one training task per executor. If you run training, you have to set
`spark.task.cpus` parameter to be equal to the number of cores in executors (`spark.executor.cores`).
This limitation might be relaxed in the future ([the corresponding issue #1622](https://github.com/catboost/catboost/issues/1622)).

Model application or feature importance evaluation do not have this limitation.