pnfs-layout · shipping · also Parallel NFS layout

pNFS layout

A pNFS layout describes how a client's file-byte range maps to storage devices for direct data access.

A layout's format and storage-access rules depend on its type. Recall asks for return; revocation withdraws access and requires enforcement at the storage devices. The two events are distinct. [1]

Linux caches layout segments separately from device-ID mappings and data-server records, with references between them. A cached segment is therefore not cached file payload. [2]

For block layouts, clients enforce file-byte ranges and allowed I/O modes because block devices lack that file-level context. Revocation fencing can cover an entire host and logical unit, affecting other layouts on the same storage units. Recall alone does not require this fencing. [1]

Relations

related

Parallel NFS (pNFS), NFS, NFSv4 stateid, NFSv4 file delegation, NFS COMMIT

Limitations

• Holding a layout is not evidence that written bytes are persistent. [3]

Citations

1. RFC 8434: Requirements for Parallel NFS Layout Types, sections 3 and 4.2; control requirements and block-layout fencing. (accessed 2026-11-09)
2. Linux kernel documentation: Reference counting in pnfs; layout segments, device mappings and data-server caches. (accessed 2026-11-09)
3. RFC 8881: NFSv4.1, sections 12.5.1, 12.5.4 and 13.7; layout access, attribute synchronization and COMMIT routing. (accessed 2026-11-09)

JSON: /data/terms/pnfs-layout.json