SNIA Developer Conference September 15-17, 2025 | Santa Clara, CA
The Windows Protocol Test Suites is an open source, cross platform application designed to enable the testing of implementations of selected Windows protocols. It is used both internally at Microsoft to ensure that protocol behaviors are in line with the publicly available protocol specifications, and by protocol implementers to test their own protocol behaviors. In this talk, we'll provide a deep dive into the architecture and design of the Test Suites, discussing some of the design decisions. We will also discuss the protocols that are currently supported, and then demonstrate using the Test Suites to test a protocol implementation.
The Windows Protocol Test Suites is an open source, cross platform application designed to enable the testing of implementations of selected Windows protocols. It is used both internally at Microsoft to ensure that protocol behaviors are in line with the publicly available protocol specifications, and by protocol implementers to test their own protocol behaviors. In this talk, we'll provide a deep dive into the architecture and design of the Test Suites, discussing some of the design decisions. We will also discuss the protocols that are currently supported, and then demonstrate using the Test Suites to test a protocol implementation.
To implement SMB2 unix extensions, smbd needs to implement ntfs reparse points to present symlinks, sockets and other special files to clients. This talk will present an overview of what reparse points are at their core and where Samba stands to implement them. Also, it will highlight the current status of the Linux kernel SMB client implementation regarding reparse points and special files over SMB2.
High-performance computing applications, web-scale storage systems, and modern enterprises increasingly have the need for a data architecture that will unify at the edge, and in data centers, and clouds. These organizations with massive-scale data requirements need the performance of a parallel file system coupled with a standards-based solution that will be easy to deploy on machines with diverse security and build environments.
Standards-Based Parallel Global File System - No Proprietary Clients
The Linux community, with contributions from Hammerspace, has developed an embedded parallel file system client as part of the NFS protocol. With NFS 4.2, standard Linux clients now can read and write directly to the storage, and scale out performance linearly for both IOPS and throughput, saturating the limits of both storage and network infrastructures. Proprietary software is no longer needed to create a high-performance parallel file system, as NFS is built into open standards and included into Linux distributions. NFS 4.2 is a commercially driven follow-on to pNFS concepts.
Today’s data architectures span multiple types of storage systems at the edge, in the data center, and in the cloud. With the rise of data orchestration systems that place data on the appropriate storage, in the optimal geographic location, NFS 4.2 is a must-have technology to deliver high-performance workflows working with distributed data sets.
Automated Data Orchestration - Across Any Storage System
Hammerspace developed and contributed the Flexible Files technology to make it possible to provide uninterrupted access to data by applications and users while orchestrating data movement even on live files across incompatible storage tiers from different vendors and multiple geographic locations.
Flexibles
Files, along with mirroring, built-in real-time performance telemetry, and attribute delegation (to name a few) are put to work in a global data environment to non-disruptively recall layouts, which enables live data access and data integrity to be maintained, even as files are moved or copied. This has enormous ramifications for enterprises as it can eliminate the downtime traditionally associated with data migrations and technology upgrades. Enterprises can combine this capability with software, such as a metadata engine, that can virtualize data across heterogeneous storage types, and automate the movement and placement of data according to IT-defined business objectives.
Building a Global Data and Storage Architecture
Hammerspace brings NFSv4.2 (in addition to SMB and NFSv3) connectivity to its parallel global file system to build a standards-based, high-performance file system that spans existing and multiple otherwise incompatible storage systems from any vendor as well as across decentralized locations. In this way it can intelligently and efficiently automate orchestration of data to applications, compute clusters, or users that need it, enabling global access for analysis, distributed workloads, or to run AI-driven insights.
With the increasing amount of network throughput, we'll reach a point where a data copies are too much for a single cpu core to handle. This talk gives an overview about how the io_uring infrastructure of the Linux kernel could be used in order to avoid copying data, as well as spreading the load between cpu cores. A prototype for this exists for quite some time and shows excellent results.
The Linux SMB3.1.1 client continues to be one of the most active filesystems in Linux with many improvements added each year, enhancing its ability to securely, reliably and efficiently access remote data. This presentation will cover new features added to the Linux client, and new features you can expect to see over the coming year. Whether accessing data from the smallest devices or the largest (and even the cloud), getting at remote files matters. The SMB3.1.1 client continues to be the most active network/cluster filesystem on Linux over the past year, with many recent enhancements, and progress on servers on Linux (Samba and also the kernel server ksmbd) has also been excellent. Over the past year, significant improvements have been made to metadata and directory caching, multichannel performance, symlink handling, remote swapfiles, improved readahead, better file caching, enhanced POSIX/Linux compatibility, improved TMPFILE support, and even improvements to DFS (global name space support). It has been a great year for the Linux client, but we will also describe some of the improvements to the Linux kernel server (ksmbd) and how the addition of the Linux kernel server has helped advance the kernel client even faster.
Current clustered Samba uses its homegrown distributed database "ctdb" as a storage backend for maintaining coherent fileserver state. ctdb predates most cloudy distributed NoSQL databases that came to rise on the wings of the likes of Google Bigtable, Amazon Dynamo, Apache Cassandra and so on in the 2000's. It has worked extremely well for the high performance scaleout NAS usecase, but the emerging shift to the cloud entails serious scalability, elasticity and managability challenges. So are there alternatives to ctdb? At SambaXP'23 I took we shared the results of a functional and performance evaluation of severl Open Source distributed databases. In this presentation I'll share the latest results of large scaleout test with FoundationDB.