Xinnor has shipped xiNAS 3.13, a production release that collapses RAID, filesystem, NFS and network administration into a single control plane, with technology previews of AI-assisted management and pNFS placement coordination close behind. The release is available now as the current stable version.

xiNAS began life earlier this year as Xinnor's move from RAID software into a fast network-attached file system. The product runs as a high-performance NFS-over-RDMA storage system on Supermicro storage servers with AMD EPYC processors and PCIe Gen5 NVMe drives.

What does the new control plane change?

Version 3.13 introduces one management workflow for storage provisioning, NFS configuration, health inspection and configuration history. Whether an administrator works from the terminal interface, CLI, REST API or an MCP client, xiNAS follows the same steps.

The system builds an end-to-end plan for the task, shows what will change, checks permissions, applies the approved operation and records it in the audit log. Operations that can alter or destroy data require the appropriate role and explicit confirmation.

The practical gain is consistency. Administrators no longer work through separate management paths for RAID, filesystems, NFS and networking. In plain terms, one audited workflow now governs the whole storage stack.

How do AI assistants fit in?

An xiNAS 3.14 technology preview extends the same control plane to MCP-compatible AI assistants and automation tools. Administrators can use an AI client to inspect system health, investigate unavailable shares, review changes or prepare configuration operations.

Sergei Platonov, VP of Strategy at Xinnor, frames the safety model directly. He said: "AI can make storage administration much easier, but it should not bypass the controls that make infrastructure predictable and safe. In the 3.14 preview, an assistant can inspect the system, explain what it finds and prepare an operation, while permissions, the execution plan and final confirmation remain under xiNAS control."

The AI client receives no shell access. Requests map to defined xiNAS operations and follow the same permissions, planning, confirmation and auditing rules as every other management interface.

Diagnostics carry their own honesty check. Results record which checks succeeded, failed, timed out or could not be performed, so missing information is never treated as evidence of a healthy system.

The takeaway is that the AI assistant proposes and explains, while xiNAS still permits, plans and confirms.

What problem does the 3.15 preview address?

A separate xiNAS 3.15 technology preview targets scale-out pNFS environments. The question it explores is how a coordination layer could account for the actual condition of storage servers when placing new data.

In the preview, xiNAS publishes the condition of the RAID, filesystem and NFS layers behind each share. A connector can feed this information to pNFS Lattice, an open-source metadata and coordination layer for pNFS.

This creates a basis for evaluating placement policies that avoid servers that are rebuilding, degraded or unavailable. Placement intelligence, in other words, could stop new data from landing on sick hardware.

Availability and roadmap

xiNAS 3.13 is available now as the current stable release. Source branches for xiNAS 3.14 and 3.15 are available separately for technical evaluation, with the 3.15 work described as an early technology preview intended for technical inspection and development.

Platonov will discuss the architecture and safety model behind the MCP preview at the SNIA Developer Conference 2026 on September 30. If the 3.14 preview holds its permission and confirmation boundaries through production, AI-assisted NAS administration will arrive without loosening the audit trail that enterprise storage buyers expect.