KNFSD NFS Cache Proxy

Executive Summary

KNFSD File Cache is described as “an open source NFS caching proxy solution running on Amazon Elastic Compute Cloud (Amazon EC2)” that extends existing NFS storage into AWS. It supports NFSv3 and NFSv4 and is positioned for high-performance, read-heavy access patterns while retaining the ability to write back to the source. The design includes a two-tier cache hierarchy per proxy instance and publishes operational telemetry to Amazon CloudWatch, with deployment packaged as a single Terraform module.

Key Industry Developments

• KNFSD File Cache is positioned as an open source NFS caching proxy that runs on Amazon EC2 to extend existing NFS servers into AWS without requiring changes to NFS clients.
• The proxy architecture includes a per-instance cache hierarchy with two tiers, enabling caching behavior across memory and local disk on each proxy instance.
• KNFSD exports a large set of operational metrics into Amazon CloudWatch, enabling monitoring based on custom telemetry emitted by the solution.
• Infrastructure provisioning is described as automated through Terraform, with scaling behavior driven by a custom CloudWatch metric tied to active NFS connections per instance.

Real-World Use Cases

• Burst-scale throughput can be added in AWS from an existing NFS source while avoiding data migration and avoiding changes to client mount paths.
• Hybrid and burst compute workflows can use a high-performance NFS caching proxy layer to provide shared access to NFS-backed datasets in the cloud.
• Distributed compute across AWS Regions or Availability Zones can use local caching to reduce latency to a central NFS source, keeping NFS access semantics while improving locality.
• Environments using Amazon EC2 Spot instances downstream can rely on the cache tier to reduce re-fetching after instance replacement, using the cache layer as a persistence point for read-heavy datasets.

Why It Matters

KNFSD’s approach targets NFS-based workflows that need cloud-side performance improvements while keeping existing NFS client behavior, since it is explicitly described as an NFS caching proxy running on Amazon EC2. The two-tier cache hierarchy provides a concrete mechanism for accelerating read-heavy access by caching data at multiple levels on each proxy instance. Operationally, the solution emphasizes observability and automation: it “publishes over 70 custom metrics to CloudWatch,” and it can be deployed via “a single Terraform module,” with scaling driven by “the number of active NFS connections per instance.” The project is also described as open source and “under the Apache 2.0 license,” which can affect how teams evaluate adoption and integration.

Sources

• AWS Media Blog — “Introducing KNFSD File Cache: Extending your NFS storage into the cloud” https://aws.amazon.com/blogs/media/introducing-knfsd-file-cache-extending-your-nfs-storage-into-the-cloud/