Zoned storage enables higher storage densities with Shingled Magnetic Recording (SMR) hard disks and reduces costs and improves tail latency with Zoned Namespace (ZNS) SSDs.

Zoned XFS adds zoned storage support to the existing scalable XFS file system for Linux to seamlessly enable the benefits of zoned storage for unmodified applications such as key-value stores or distributed storage systems with performance beating that of existing zone-aware file systems on Linux.

We present an overview of Zoned XFS design and performance evaluation results which show that Zoned XFS achieves better throughput with SMR hard disks for small and large file read and write workloads compared to the BTRFS and F2FS file systems, even in the presence of garbage collection with a nearly full file system. With RocksDB, we show that Zoned XFS on a ZNS SSD achieves higher read throughput with mixed read/write workloads and overwrite workloads compared to F2FS. Furthermore, Zoned XFS compares favorably to the RocksDB special purpose ZenFS zoned storage backend.