# Kafka performance: IO

Published 2022-12-03

This tutorial explains to you the impact of the IO on the Kafka server performance.

Here are some information to retain about the impact of the disks on Kafka cluster performance:

  • Disks (IO) could be a bottleneck in Kafka : there are a lot of read/write operation on disks.

  • The write/read operations are done sequentially and asynchronous so, the disk must be optimized for sequential operations.

  • a SSD disk don’t deliver a big advantage due to Kafka’s sequential disk I/O paradigm.

  • If you configure multiple data directories, the broker places a new partition in the path with the least number of partitions currently stored. Each partition will be entirely in one of the data directories. If data is not well balanced among partitions, this can lead to load imbalance among disks.

  • Avoid network-attached storage (NAS). NAS is often slower, displays larger latencies with a wider deviation in average latency, and is a single point of failure.

  • Format your drive as XFS : no tuning required.

  • Monitor constantly the performance and the free space on Kafka broker disks.