Loading…
This event has ended. Create your own event → Check it out
This event has ended. Create your own
View analytic

Sign up or log in to bookmark your favorites and sync them to your phone or calendar.

Wednesday, March 11
 

9:00am

The Future for Linux Storage, Filesystems and Memory Management - James Bottomley (CTO, Server Virtualization at Parallels, Jeff Layton (Senior Software Engineer at Primary Data) and Rik van Riel (Principal Software Engineer at Red Hat)
The three track leaders from the recent LSF/MM Summit will present the outcomes from the summit and some indications of the future directions the Linux ecosystem is taking.

Wednesday March 11, 2015 9:00am - 9:40am
Space 57

9:40am

At-Scale Datacenters and the Demand for New Storage Architectures - Allen Samuels, Chief Software Architect, Emerging Storage Solutions at SanDisk
Join Allen as he discusses at-scale datacenters and the demand for new storage architectures.

Speakers
avatar for Allen Samuels

Allen Samuels

Engineer Fellow, Western Digital
Allen joined SanDisk in 2013 as an Engineering Fellow, he is responsible for directing software development for SanDisk’s system level products. He has previously served as Chief Architect at Weitek Corp. and Citrix, and founded several companies including AMKAR Consulting, Orbital Data Corporation, and Cirtas Systems. Allen has a Bachelor of Science in Electrical Engineering from Rice University.


Wednesday March 11, 2015 9:40am - 10:00am
Space 57

10:00am

Break
Wednesday March 11, 2015 10:00am - 10:30am
Foyer

10:30am

Btrfs: Future Plans - Chris Mason, Facebook
As the Btrfs filesystem matures, we are rounding out an impressive list of features. This talk will describe what make Btrfs unique, layout our roadmap for the next year, and demonstrate some of the powerful new features we have in development.

Speakers
CM

Chris Mason

Software Engineer, Facebook
Chris is a Software Engineer on the kernel team at Facebook, and the maintainer of the Btrfs filesystem. He has been working full time on the kernel for over 16 years, and lives in Rochester New York.


Wednesday March 11, 2015 10:30am - 11:20am
Loft 2

10:30am

Ceph Performance Demystified: Benchmarks, Tools, and the Metrics that Matter - Adolfo Brandes, Hastexo
The Ceph distributed storage platform is a complex array of subsystems, and its overall performance is influenced by factors as diverse as local block storage IOPS and streaming writes, network throughput and latency, CPU utilization on servers and clients, Ceph configuration, and many other factors. When optimizing Ceph for performance, the many variables involved can make the task daunting. Fortunately, though, the intrepid storage administrator needs only a handful of tools to quickly establish performance metrics at all levels of the stack — and some of these tools are even built right into Ceph.

This presentation covers important Ceph performance metrics and tools, provides a walkthrough of bottom-up Ceph benchmarking, and provides guidance on how to properly interpret results and tune for performance optimization.

Speakers
avatar for Adolfo Brandes

Adolfo Brandes

Senior Consultant, hastexo
Adolfo started out as an ESL teacher, but a gift for programming computers shaped his early career. He developed everything from an Asterisk-based PBX to the odd Linux kernel driver, including several websites and an AGPL3 Python-backed web game. Nevertheless, in his mid-thirties Adolfo saw the opportunity to unite two talents, and once more turned to mentoring by teaching developers how to contribute to OpenStack more efficiently. Most recently... Read More →


Wednesday March 11, 2015 10:30am - 11:20am
Carver 1

10:30am

Evaluation of Filesystems for Shingled Drives - Hannes Reinecke, SUSE
Shingled drives impose some limitations on the I/O patterns. Especially the write path requires a strict sequential ordering. Normally filesystems or the block layer are not imposing any ordering requirements, so it'll depend on the filesystem implementation if the ordering can be guaranteed. In this presentation I'll evaluate current general-purpose filesystems (btrfs, ext4, and xfs) and a dedicated filesystem (nilfs) for this ordering guarantee and hence their suitability for shingled drives. Additionally I'll present strategies for deploying filesystems on shingled drives.

Speakers
avatar for Hannes Reinecke

Hannes Reinecke

Team lead Storage & Networking, SUSE Linux GmbH
Studied Physics with main focus image processing in Heidelberg from 1990 until 1997, followed by a PhD in Edinburgh 's Heriot-Watt | University in 2000. | Now working at SUSE Labs with focus on storage and mainframe. Principal contact point for storage related issues on SLES. | Currently I'm working on dusting out murky corners of the linux SCSI stack, and trying to get SMR drives to ... well, not exactly fly, but at least moving at a... Read More →


Wednesday March 11, 2015 10:30am - 11:20am
Studio 3

10:30am

Year of the Linux I/O Cache Hinting Solution - Dan Williams, Intel
While it has yet to claim as many years of expected imminent arrival, an I/O cache hinting solution for Linux has been under serious consideration since at least LSF 2012. All the major storage interface standards bodies (NFS, T10, T13, and NVME working groups) have each defined cache attributes that can be specified in their respective transports. Perhaps the largest hurdle to overcome for this effort is finding the minimal set of hints that unlocks the bulk of the potential of capable devices without committing the Linux kernel to maintaining an overly complicated hinting framework. This presentation seeks to stand on the foundation of Martin Petersen’s work in this space (wrangling device hinting schemes vs fadvise() in the kernel) and reconcile it with learnings from the Intel effort to implement support for T13/SATA-IO hints in current Solid State Hybrid Drives (SSHDs).

Speakers
DW

Dan Williams

Intel
Dan is a Linux kernel developer in Intel Open Source Technology Center. He primarily works on enabling platform storage technologies. Most recently he has been involved in persistent memory enabling as a maintainer of the Linux kernel libnvdimm sub-system. He led the Persistent Memory micro-conference at Linux Plumbers 2015 and presented at the inaugural Vault Conference.


Wednesday March 11, 2015 10:30am - 11:20am
Carver 2

11:30am

A Simple, and Scalable pNFS Server For Linux - Christoph Hellwig
The parallel NFS extensions in NFSv4.1 have seen far less deployment than the hype around them would suggest.  One big reason for that is few usefulservers have been available, and the other one are over complicated specification that are almost unimplementable.

This presentations explains the high level concepts of pNFS, and their heritage.  It continues with a detailed explanation of a Linux Kernel implementation of the pNFS block layout server, which has been kept very simple and small despite the protocol challenges.

Speakers
CH

Christoph Hellwig

Christoph Hellwig has been working with and on Linux for the last ten years, dealing with kernel-related issues much of the time. In addition he is or was involved with various other Open Source projects. After a number of smaller network administration and programming contracts he worked for Caldera's German development subsidiary on various kernel and userlevel aspects of the OpenLinux distribution. Since 2004 he has been running his own... Read More →


Wednesday March 11, 2015 11:30am - 12:20pm
Loft 2

11:30am

Ceph FS Development Update - John Spray, Red Hat
Ceph provides a resilient and scalable storage model (RADOS) using clusters of commodity hardware. Along with the RADOS block device (RBD), and the RADOS object gateway (RGW), Ceph also provides a POSIX filesystem interface (Ceph FS). While RBD and RGW have been in use for production workloads for some time, efforts to make Ceph FS ready for production are now underway.

This presentation will introduce Ceph FS, explain its architecture, and provide an update on the latest development and testing work being done improve stability and manageability of the filesystem.

Speakers
avatar for John Spray

John Spray

Red Hat
John is a Ceph developer at Red Hat, working from his home office in Edinburgh, Scotland. He works mainly on the Ceph filesystem layer (CephFS).


Wednesday March 11, 2015 11:30am - 12:20pm
Carver 1

11:30am

Modelling and Estimating Distributed Storage Characteristics - Lars Marowsky-Brée, SUSE
Distributed software-defined storage (such as Ceph) provides for scalable and flexible storage systems, covering the full range from cold and long-term data retention, general purpose storage, cloud storage, big data to high performance compute applications. Compared to legacy appliances, the full set of configuration and architecture choices at all layers (server, network, software) is exposed. In theory, this allows them to be infinitely customizable. In practice, their performance characteristics are non-trivial to predict and estimate without a costly large pilot.

Here, we will discuss relevant design choices and their trade-offs, and how they will affect the performance and cost of the resulting system. Research into a model and simulation of actual Ceph deployments is also included.

Speakers
avatar for Lars Marowsky-Brée

Lars Marowsky-Brée

Distinguished Engineer, SUSE
Lars serves as architect for Software-Defined-Storage at SUSE, focused on Ceph. He represents SUSE on the Ceph Advisory Board as a founding member, and is a frequent invited speaker. Since joining SUSE in 2000, his previous roles include senior consultant, kernel engineer, engineering manager in SUSE Labs, and he is most known for his work on the Linux High-Availability stack. He was named Distinguished Engineer in 2012. He holds an Msc from... Read More →


Wednesday March 11, 2015 11:30am - 12:20pm
Carver 2

11:30am

SMR in Linux Systems -- Seagate's Contribution to Legacy Filesystems - Adrian Palmer, Seagate
Seagate’s Shingled Magnetic Recording Friendly File System (SMRFFS) is a effort to enable SMR use in Linux systems by taking advantage of the new zoned block device interfaces standards. Current developmental work focuses on ext4, with xfs and others to follow as SMR drives begin to permeate the market. Changes in ext4 includes modifications that respect the primary rule introduced by SMR: forward-write preferred. We demonstrate taking the T10/13 specs and aligning them in the legacy File System. This reduces the time for market convergence and the continued use of legacy software on upgraded legacy filesystems with new SMR drives. By transforming ext4 into a Host Aware SMR Friendly File System, we leverage the reliability record and performance of ext4 with a new generation of hardware.

Speakers
AP

Adrian Palmer

Adrian Palmer holds a MS of Electrical Engineering and a BS of Computer Engineering and Math, all from the University of Wyoming. He has spent over 10 years in Systems Administration, and has hobbyist experience in Linux and UNIX systems, creating router/gateway and NAS systems. In 1999, he was the first public high school student to earn an MCSE. Adrian currently works at Seagate and leads the SMRFFS project, contributing to the open source... Read More →


Wednesday March 11, 2015 11:30am - 12:20pm
Studio 3

12:20pm

Lunch
Wednesday March 11, 2015 12:20pm - 2:00pm
TBA

2:00pm

Deploying Btrfs in Production - Josef Bacik, Facebook
Btrfs has been in development for almost 8 years now and has finally become mature enough to be deployed in production. In this presentation I will talk about the challenges we have faced at Facebook with our various deployments of Btrfs and the tools we have used to track down problems. I will talk about our efforts around testing and validation, including a robust power fail testing framework. Facebook has hundreds of thousands of machines that are divided among many different workloads, each of these environments provide their own challenges with interesting solutions for Btrfs.

Speakers
avatar for Josef Bacik

Josef Bacik

Software Engineer, Facebook
Josef is a software engineer at Facebook working on Btrfs and helping to maintain the Facebook kernel. He has been working on Btrfs since the beginning and is one of the maintainers. 


Wednesday March 11, 2015 2:00pm - 2:50pm
Studio 3

2:00pm

Distributed Storage and Compute With Ceph's librados - Sage Weil, Red Hat
The Ceph distributed storage system sports object, block, and file interfaces to a single storage cluster. These interface are built on a distributed object storage and compute platform called RADOS, which exports a conceptually simple yet powerful interface for storing and processing large amounts of data and is well-suited for backing web-scale applications and data analytics. In features a rich object model, efficient key/value storage, atomic transactions (including efficient compare-and-swap semantics), object cloning and other primitives for supporting snapshots, simple inter-client communication and coordination (ala Zookeeper), and the ability to extend the object interface using arbitrary code executed on the storage node. This talk will focus on librados API, how it is used, the security model, and some examples of RADOS classes implementing interesting functionality.

Speakers
avatar for Sage Weil

Sage Weil

Ceph Principal Architect, Red Hat
Sage Weil originally helped design and build Ceph as part of his graduate research at UC Santa Cruz. Since then, he has continued to lead the engineering effort, building Ceph into an open source, unified storage platform with an emphasis on scalability, reliability, and performance for cloud workloads. He currently works as Red Hat in the Office of Technology, where his primary responsibility is to lead the Ceph development community while... Read More →


Wednesday March 11, 2015 2:00pm - 2:50pm
Carver 1

2:00pm

Increasing SCSI LLD Driver Performance by Using the SCSI Multiqueue Approach - Bart Van Assche, SanDisk
Using the full potential of all-flash arrays not only requires a high-bandwidth and low-latency communication network but also requires optimized SCSI LLD drivers. The SCSI multiqueue approach that has been merged in Linux kernel 3.17 enables higher performance and lower latency. Levering the full potential of the SCSI multiqueue approach is only possible by integrating support for multiple hardware queues in the SCSI LLD. The SCSI RDMA Protocol (SRP) initiator driver is a SCSI LLD driver that enables access to remote storage over network technologies that support RDMA, e.g. InfiniBand, iWARP and RoCE. During this presentation it will be explained why the recently posted SRP initiator patches resulted in a very significant performance improvement and also how that performance improvement has been realized.

Speakers
BV

Bart Van Assche

Bart Van Assche obtained a Ph.D. in distributed computing in Belgium and lives in Belgium. He is the maintainer of the SRP initiator driver in the Linux kernel and is also an active contributor to the Linux kernel SCSI core. Bart works for SanDisk on the ION Accelerator software. This software provides SAN functionality that is optimized for flash memory and that has already established several performance world records.


Wednesday March 11, 2015 2:00pm - 2:50pm
Loft 2

2:00pm

LIO and the TCMU Userspace Passthrough: The Best of Both Worlds - Andy Grover, Red Hat
The LIO Kernel Target has given us enhanced stability, performance, and features compared to previous daemon-based iSCSI targets. Its support for multiple transport fabrics enables support for SCSI transports besides iSCSI, and simplifies target configuration for all fabrics.

However on the backstore side, being 100% kernel code can be limiting -- at least until the recent addition of the "TCMU" backstore. TCMU allows LIO to delegate data storage to userspace processes, enabling richer options for file formats, clustering, and also experimentation.

Andy will discuss the implementation of the TCMU backstore and its kernel-to-user communication method, and ongoing feature and performance work. There will also be information on how interested developers can get started working with TCMU, and enable new data representations to be exported over multiple SCSI fabrics by LIO.

Speakers
AG

andy grover

Red Hat
Andy Grover is a Principal Software Engineer at Red Hat, working to improve Linux’s block storage capabilities. His work encompasses both changes to the kernel itself as well as low-level management tools. Previous work areas include networking and ACPI. He has previously spoken at Plumbers, Linux Kernel Summit, OSCON, LinuxCon, and Open Source Bridge, among others. He lives in Portland.


Wednesday March 11, 2015 2:00pm - 2:50pm
Carver 2

3:00pm

Cephfs fsck: Distributed File System Checking - Gregory Farnum, Red Hat
Red Hat is trying to move the CephFS distributed file system from a community project to something commercially supportable. As part of that, the CephFS team has been working on an fsck solution to verify integrity and correct any unexpected errors from either physical catastrophe or software error in CephFS itself or the underlying RADOS/Linux stack. Distributed and scalable fsck built on top of an object store interface is very different from the traditional fsck on top of a block interface to a single disk, so this talk will discuss the overall vision, current status, and future work of our solution.

Speakers
avatar for Gregory Farnum

Gregory Farnum

Principal Software Engineer, Red Hat
Greg Farnum is a long-standing member of the core Ceph development group, having joined the project as the third full-time engineer after graduating from Harvey Mudd College in 2009. A Red Hat employee, Greg has previously served many development roles throughout the Ceph project and currently acts as the filesystem team technical lead.


Wednesday March 11, 2015 3:00pm - 3:50pm
Carver 1

3:00pm

Ext4 Filesystem Scaling - Jan Kára, SUSE
Ext4 is one of the mainstream file systems used in Linux. In this presentation, Jan Kara will talk about some of the work he and other ext4 developers did to improve scalability and performance of the file system. Presented topics will include reducing contention resulting from handling of list of inodes with truncate in progress (called orphan list), improving latency of shrinking of a cache of logical-to-physical block mappings, and more.

Speakers
JK

Jan Kara

SUSE Labs, SUSE
Jan Kara is doing Linux kernel hacking in file systems area over 15 years. He is the maintainer of ext3 file system, udf file system, and quota subsystem, working also on other file systems, writeback logic, notification framework, and other miscellaneous stuff. Currently he is working as a kernel engineer for SUSE. | | The speaker led several sessions in past Kernel Summits and Linux Storage, Filesystem, and MM Summits.


Wednesday March 11, 2015 3:00pm - 3:50pm
Loft 2

3:00pm

Lambert: Achieve High Durability, Low Cost & Flexibility at Same Time — open source cold storage engine for ExaBytes data in Alibaba - Robin Dong, Alibaba & Coly Li (Alibaba)
Tape or optical disc storage system might not always be perfect solutions for cold data storage, especially for cloud computing service providers, when there is no standard data center to deploy large scaled automatic mechanical equipments.

Alibaba coups with a similar situation in China for years. Before having enough own-built data centers, to store rapidly expended cold data in many middle-size and non-standard data centers, tape or optical disc solution is not the proper answer. To achieve high durability, low cost and flexibility at the same time for cold data storage, one of solutions from Alibaba is to build cold storage engine by cheap, low performance commercial hardware and simple open source software. We call this cold storage engine “Lambert” as code name, it simply runs sheepdog (an open source project for distributted storage) to manage huge number of SMR hard disks.

Speakers
avatar for Robin Dong

Robin Dong

Developer and architect for the distributed cold storage system. Graduated from Peking University in 2006, with a master degree in Software Development. Joined Alibaba Group after graduated, first as a developer for low-level common library , and then joined Alibaba Linernel team at 2011 and working on Ext4, Flashcache, IO scheduler.
CL

Coly Li

Software engineer in Alibaba (NYSE: BABA) infrastructure development team, he is the first engineer in Alibaba Linux kernel team and help to build the kernel team. Before joined Alibaba, he was Linux kernel C developer in SuSE Labs working on OCFS2, Ext4, YaST and OpenIPMI projects.


Wednesday March 11, 2015 3:00pm - 3:50pm
Carver 2

3:00pm

NFSv4.2 and Beyond - J. Bruce Fields, Red Hat
NFSv4 is now default in recent linux distributions and provides a platform for further protocol extensions. Among other features that have recently arrived or are in progress:
  • improvements useful for storage of VM images, including sparse file support, server-side copy, and labeled (selinux-enabled) NFS.
  • features which abstract the location of data, such as pNFS and fedfs.
We'll also discuss the protocol development process and possible future directions.

Speakers
JB

J. Bruce Fields

J. Bruce Fields is a developer on Red Hat's Red Hat NFS team, the maintainer of the Linux kernel NFS server, a contributor to the IETF's NFSv4 working group, and was previously a member of the University of Michigan team that wrote the first NFSv4 Linux implementation.


Wednesday March 11, 2015 3:00pm - 3:50pm
Studio 3

3:50pm

Break
Wednesday March 11, 2015 3:50pm - 4:20pm
Foyer

4:20pm

Challenges of providing Remote DMA access to Persistent Memory - Matthew Wilcox, Intel
Persistent Memory poses some new challenges to the Linux kernel. One of the trickiest problems is simultaneously supporting direct access to persistent memory and Remote DMA access. The DAX infrastructure provides direct access to persistent memory by removing the page cache, but the page cache provides an important layer of indirection between the filesystem storage and the memory used to cache that file.

This talk will present three different solutions to the problem and contrast their effects on filesystems, the VM, the VFS and RDMA device drivers.

Speakers
MW

Matthew Wilcox

Matthew works for Microsoft on Persistent Memory. He has presented at many Linux conferences, including Vault, LinuxCon, LCA and OLS. His most recent projects include NVM Express and USB Attached SCSI.


Wednesday March 11, 2015 4:20pm - 5:10pm
Studio 3

4:20pm

Optimizing Ceph Performance For All-Flash Architectures - Viyayendra Shamanna, SanDisk
Ceph is a massively scalable, open source, software-defined distributed storage stack, which uniquely provides object, block and file access interfaces in one unified system. Most of the current Ceph deployments use SSDs for Ceph journal and hard drives as Ceph data store. But when the hard disk backend was replaced by an all-SSD infrastructure, multiple CPU bottlenecks in Ceph code were exposed. Major improvements were done to the Ceph OSD code to eliminate these bottlenecks. Some of the notable changes were removing a single dispatch queue bottleneck, a new sharded worker thread pool for efficient request processing, improved lock granularity, sharded file descriptor cache etc which resulted in performance improvements of 6X for block reads and 5X for object reads.

Speakers
VS

Vijayendra Shamanna (“Viju”)

Vijayendra Shamanna (“Viju”) is a Senior Manager with the Emerging Storage Solutions software team at SanDisk, Bangalore. SanDisk is a global leader in flash storage solutions from edge devices to cloud and enterprise data centers. Viju has over 18 years of experience as Senior Engineering Manager, Architect and Principal Engineer, delivering complex Storage, Virtualization, Networking and Data Center solutions from inception to customer... Read More →


Wednesday March 11, 2015 4:20pm - 5:10pm
Carver 1

4:20pm

Reducing File System Tail Latencies with Chopper - Jun He, University of Wisconsin-Madison
Jun He (UW-Madison) will present Chopper, a tool that efficiently explores the vast input space of file system policies to find behaviors that lead to high-tail latencies. Jun He will focus specifically on block allocation, as complex allocation policies can lead to surprising and costly performance problems. The approach utilizes sophisticated statistical methodologies to explore the search space efficiently while still ensuring excellent coverage of system responses. Chopper has been applied to study the over all behavior of two file systems, and to study Linux ext4 in depth. Chopper is able to identify four internal design issues in ext4 which form a large tail in its performance distribution. By removing the underlying problems in the code, consistent and satisfactory file layouts are achieved.

Speakers
avatar for Jun He

Jun He

PhD Student, University of Wisconsin-Madison
Jun He is a PhD student at Department of Computer Sciences, University of Wisconsin-Madison. He works with Prof. Andrea C. Arpaci-Dusseau and Prof. Remzi H. Arpaci-Dusseau. Recently, he has become interested in taking advantage of statistics techniques, especially the Design Of Experiments (DOE) and Sensitivity Analysis, to guide computer system designing, which could help developers to avoid unexpected behaviors and make more confident... Read More →


Wednesday March 11, 2015 4:20pm - 5:10pm
Loft 2

4:20pm

Skylight: A Window on Shingled Disk Operation - Abutalib Aghayev, Northeastern University Solid State Storage Lab
We introduce Skylight, a novel methodology that combines software and hardware techniques to reverse engineer key properties of drive-managed SMR. The software component of Skylight measures latency of controlled I/O operations to infer important properties of drive-managed SMR, including type, structure, and size of the persistent cache, type of cleaning algorithm, type of block mapping, and size of shingled bands.
The hardware component of Skylight tracks drive head movements, using a high-speed camera through an observation window drilled through the cover of the drive. These observations confirm inferences from measurements, and resolve ambiguities which arise from use of performance measurements alone. We show the generality and efficacy of our techniques by running them on top of emulated and real SMR drives, discovering valuable performance-relevant details.

Speakers
AA

Abutalib Aghayev

PhD Student, Northeastern University
I am a graduate student at Northeastern University Solid State Storage Lab doing research in storage with Prof. Peter Desnoyers. I recently finished characterization work for SMR (Shingled Magnetic Recording) drives and now working on a follow-up project researching file system optimizations for SMR drives and alternative STL (Shingle Translation Layer) designs.


Wednesday March 11, 2015 4:20pm - 5:10pm
Carver 2

5:10pm

Booth Crawl & Happy Hour
Join fellow attendees Wednesday after conference sessions conclude for a booth crawl including drinks and appetizers.  

Wednesday March 11, 2015 5:10pm - 7:00pm

7:00pm

BoFs: NFS - Christoph Hellwig
Discussion on Linux NFS client and server implementation issues, with a focus on new features in NFSv4.2 and pNFS.

Moderators
CH

Christoph Hellwig

Christoph Hellwig has been working with and on Linux for the last ten years, dealing with kernel-related issues much of the time. In addition he is or was involved with various other Open Source projects. After a number of smaller network administration and programming contracts he worked for Caldera's German development subsidiary on various kernel and userlevel aspects of the OpenLinux distribution. Since 2004 he has been running his own... Read More →

Wednesday March 11, 2015 7:00pm - 8:00pm
Loft 2

7:00pm

BoFs: SMR and Zoned Block Devices: Seagate & HGST - Adrian Palmer, Seagate & Jorge Campello, HGST
Come and discuss the issues & proposed solutions of enabling the benefits of SMR-based HDDs in Linux.  Engineers & developers from HGST and Seagate will be on hand to explain the planned approaches, answer questions, listen to feedback, and describe the future of ZBD (zoned block devices).

Moderators
avatar for Jorge Campello

Jorge Campello

Global Director of Systems and Solutions, HGST
Jorge Campello is the Global Director of Systems and Solutions at HGST Research. He received his PhD in Electrical Engineering from Stanford University in 1999 and has been working in the storage industry over the last 16 years. An SMR evangelist, he was deeply involved in the formation of the standardization efforts around SMR and now directs his efforts at the libzbc open source project. He is also Co-Chair of the Storage Work Group within... Read More →
AP

Adrian Palmer

Adrian Palmer holds a MS of Electrical Engineering and a BS of Computer Engineering and Math, all from the University of Wyoming. He has spent over 10 years in Systems Administration, and has hobbyist experience in Linux and UNIX systems, creating router/gateway and NAS systems. In 1999, he was the first public high school student to earn an MCSE. Adrian currently works at Seagate and leads the SMRFFS project, contributing to the open source... Read More →

Wednesday March 11, 2015 7:00pm - 8:00pm
Carver 1

7:00pm

BoFs: The Future of Gluster - Dan Lambright, Red Hat
Come and discuss upcoming Gluster features, including erasure codes, cache tiers, NFS Ganesha, and scalability enhancements. Several developers will be on hand to answer questions, hear suggestions, and give a peek into whats next with this open source scale-out file system.

Speakers
avatar for Dan Lambright

Dan Lambright

Software Engineer, Red Hat
Dan Lambright is a principal software engineer at Red Hat, where he works on distributed storage systems. Prior to Red Hat is worked at EMC, DELL, and several storage startups. He also teaches as an adjunct professor at the University of Massachusetts, Lowell.


Wednesday March 11, 2015 7:00pm - 8:00pm
Studio 3
 
Thursday, March 12
 

9:00am

Filesystem Fuzzing Simplified - Sasha Levin, Oracle
Filesystem testing has been mostly focused around using "standard" testcase based tools. While they provide good coverage and are good for testing for known regressions, they miss quite a lot.

I'd like to present an approach which allows fuzz testing a filesystem by loading the target filesystem into a disposable harness built around the kvm tool and trinity with the ability to use the traditional test tools in parallel.

This approach, when targeted at specific subsystems has produced interesting results:
- CVE-2014-8086 (EXT4 DoS)
- CVE-2014-8559 (FS remote DoS)
- CVE-2014-4171 (shmem DoS)
- CVE-2014-3940 (procfs DoS)
- CVE-2013-7348 (AIO memory corruption)

And quite a few more issues that are just plain bugs.

Speakers
SL

Sasha Levin

Verizon Labs
Sasha is the maintainer of the 3.18 and 4.1 stable trees. He is also the maintainer of the linux-stable-security project which provides critical security updates to projects that use stable-like trees.nnSasha is currently employed by Oracle, working in the Ksplice group. We provide rebootless security updates to the Linux kernel without requiring a reboot. Previously he worked in Host Dynamics which provided the ability to dynamically shape the... Read More →


Thursday March 12, 2015 9:00am - 9:50am
Carver 1

9:00am

Open-Channel Solid State Drives - Matias Bjørling, University of Copenhagen
A new type of SSDs is under development, referred to as Open-Channel SSDs. These SSDs expose additional information about its geometry and enables the host to make internal data placement decisions on their behalf. By moving placement logic to the host, the design space for alternative storage interfaces that go beyond the traditional block device interface opens up. Examples of these interfaces include flash-optimized file-systems, and key-value stores.

This presentation will give an introduction to Open-Channels SSDs and the kernel integration, i.e., LightNVM. The main objective is to present this new storage interface and report on its support in the kernel. It will outline the open-channel SSD interface and abstraction model, the LightNVM kernel implementation, and the present two common targets: generic block interface and direct flash block access using a key-value interface.

Speakers
avatar for Matias Bjørling

Matias Bjørling

Member of Technical Staff, CNEX Labs
Matias Bjørling obtained a Ph.D. in operating systems and Solid State Drives from IT University of Copenhagen. He is currently a member of CNEX Labs technical staff. Before that, he worked with performance characterization of flash-based SSDs, operating- and database systems research and the multi-queue block layer. Matias is a contributor to the Linux Kernel and is the maintainer of the LightNVM subsystem for Open-Channel SSDs.


Thursday March 12, 2015 9:00am - 9:50am
Carver 2

9:00am

Optimizing FUSE for Cloud Storage - Maxim Patlasov, Parallals
Filesystem in Userspace (FUSE) framework has notorious reputation of a toy filesystem which is mostly used by students for educational purposes. In this presentation, Maxim Patlasov will review the challenges a company faces trying to implement enterprise-quality distributed storage solution based on FUSE framework, discuss FUSE improvements made to overcome discovered limitations (asynchronous handling of direct I/O, writeback cache policy, and others), and illuminate the areas of future development.

The work was done to support our own Parallels Cloud Storage product, but the FUSE enhancements are in the upstream kernel and available to any cloud storage system using fuse (like Gluster).

Speakers
MP

Maxim Patlasov

Lead Developer, Parallels
Graduated from MSU. Worked for Data Foundation, Jet Infosystems, Cluster File Systems, Sun Microsystems, Oracle


Thursday March 12, 2015 9:00am - 9:50am
Studio 3

9:00am

Putting the Pieces Together for a 100% F/OSS Storage Array? - Andy Grover, Red Hat
What is the "special sauce" that enables vendors to charge huge premiums for an enterprise-class storage array? These days we have many of the pieces -- LVM, raid, scalable filesystems, SAN and NAS target capability -- and we even have the beginnings of a remote API solution with libstoragemgmt and targetd. It seems like more than one vendor's value-add consists solely of an attractive and colorful web UI! Not that some aren't innovating, but isn't it about time F/OSS commoditizes what they were shipping ten years ago? Is this something the F/OSS community should aim for, or is the attraction for customers really the support, and the warm feeling of a product from a known, commercial entity?

Speakers
AG

andy grover

Red Hat
Andy Grover is a Principal Software Engineer at Red Hat, working to improve Linux’s block storage capabilities. His work encompasses both changes to the kernel itself as well as low-level management tools. Previous work areas include networking and ACPI. He has previously spoken at Plumbers, Linux Kernel Summit, OSCON, LinuxCon, and Open Source Bridge, among others. He lives in Portland.


Thursday March 12, 2015 9:00am - 9:50am
Loft 2

10:00am

DRBD9 - High Performing Reliable Block Storage for the Cloud - Philipp Reisner, LINBIT
DRBD9 is a Linux block device driver, that allows to mirror Linux block devices to multiple peer hosts. It can operate in synchronous or asynchronous mode.
Drbdmanage is a management solution, that takes over management of logic volumes (LVM) and management of configuration files for DRBD. It offers commands to create replicated volumes, take snapshots, resize, remove, etc... It allows to create volumes in consistency groups.
These capabilities are available on the command line, and as a D-BUS service. They are useful as they are, since it automates the operation of DRBD replicated storage.
Originally DRBD was mainly used to build highly available server clusters. Drbdmanage not only eases the use of DRBD in this area it opens a new class of use cases for DRBD in the field of highly automated cloud deployments. A cinder (OpenStack) driver for the mentioned stack is in development.

Speakers
avatar for Philipp Reisner

Philipp Reisner

CEO, LINBIT
Philipp Reisner is the CTO of LINBIT HA-Solutions Vienna. During his studies of Computer Science at the Technical University in Vienna (TU Wien), he started work on DRBD® replication software, which quickly gained popularity around the globe. DRBD was later accepted into mainline Linux with the 2.6.33 release. Philipp is an internationally renowned OSS specialist, kernel programmer and eminent lecturer on High Availability at international Linux... Read More →


Thursday March 12, 2015 10:00am - 10:50am
Carver 1

10:00am

GlusterFS - Overview & Future Directions - Vijay Bellur, Red Hat
GlusterFS is a general purpose distributed scale-out filesystem that runs on commodity hardware. In this presentation, Vijay Bellur will provide an architectural overview of GlusterFS and discuss how this architecture can be used to build a scale-out storage solution for modern datacenter needs. Details on new features in GlusterFS 3.6 & 3.7 releases , use cases and interesting challenges with GlusterFS will be provided. As part of this session, Vijay will also discuss integration of GlusterFS with other open source ecosystems like OpenStack, Apache Hadoop, oVirt and provide future directions of the GlusterFS project including GlusterFS 4.0.

Speakers
VB

Vijay Bellur

Co-maintainer for the GlusterFS Project, Red Hat
Vijay Bellur is a co-maintainer for the upstream GlusterFS project and was an architect at Gluster before its acquisition by Red Hat in 2011. He has been involved with building enterprise storage and scalable, distributed systems for the past decade. Vijay works out of the Red Hat office in Boston and his topics of interest include file systems, cloud technologies, and Big Data.


Thursday March 12, 2015 10:00am - 10:50am
Carver 2

10:00am

Open Source, Scale-out, Clustered NAS using nfs-ganesha and GlusterFS - Anand Subramanian, Red Hat
In this session we will discuss some important aspects in integrating nfs-ganesha, a user-space, modular and extensible NFS server with the scale-out, distributed GlusterFS filesystem to create an effective Clustered NAS solution. The integration of nfs-ganesha with GlusterFS enables access to the filesystem using NFSv3, NFSv4/4.1 and we will also look at an Open Source pNFS server implementation using nfs-ganesha and GlusterFS. We will also look at the salient features of nfs-ganesha making it a great fit for a scale-out, Clustered NAS environment like the modular FSAL (File System Abstraction Layer) architecture and the CMAL (Cluster Manager Abstraction Layer) allowing easy integration with any Cluster Manager, Cluster-wide lock manager and lock recovery aspects etc. Finally we will also try to cover some of the important use-cases of nfs-ganesha.

Speakers
AS

Anand Subramanian

Anand is a Senior Principal Engineer working in the Red Hat Storage Server product and the GlusterFS and nfs-ganesha communities. He leads the nfs-ganesha work at Red Hat and the GlusterFS-Ganesha integration. Anand has largely worked on large-scale distributed infrastructure, Clustering and Highly Available systems. He currently leads the active-active multi-headed NFS implementation using nfs-ganesha and glusterfs as well as a pNFS server... Read More →


Thursday March 12, 2015 10:00am - 10:50am
Loft 2

10:50am

Break
Thursday March 12, 2015 10:50am - 11:20am
Foyer

11:20am

Architecture and Development of OrangeFS Version 3 - Walter B. Ligon III, Clemson University
This presentation will start with a brief overview of the OrangeFS file system. Then, diving into the architecture of OrangeFS v3, which is under heavy development. Discussions will include upcoming architectural changes and major advancements in the areas of availability, security, data integrity, maintenance and administration and monitoring. Topics will include distributed replicated file and metadata objects, policy based object location, and parallel background jobs.

Speakers
WB

Walter B. Ligon III

Professor of Computer Engineering, Clemson University
A Professor of Computer Engineering, Walt has worked in parallel I/O for the past 20 years. He was a founding architect of PVFS and now now leads design and development efforts for OrangeFS, while engaging his students, collaborating with the development resources of Omnibond Systems and leveraging his associations with a number of national labs. He has a Ph.D. in Computer Science from the Georgia Institute of Technology.


Thursday March 12, 2015 11:20am - 12:10pm
Carver 2

11:20am

Asynchronous Samba - Jeremy Allison, Google
The Samba project code is moving towards an asynchronous model (with threads hiding under the covers) to improve scalability and performance. Learn how and why this transition is being done, and what effects this has on implementing the advanced features of the SMB3.1 protocol inside Samba.

This talk is for developers who use Samba to create storage products around SMB3.x, testers who work on Samba-based storage products, and storage system architects who need to tune storage products for maximum performance. The spin-off libraries created by Samba that are also used in other Open Source projects (such as sssd) will also be discussed, allowing developers to evaluate their usefulness in their own projects.

Speakers
JA

Jeremy Allison

Google
Jeremy Allison is the co-creator of the Samba project and a lead developer on the Samba Team, a group of programmers developing a Free Software Windows compatible authentication, file and print server product for Linux and UNIX systems. Samba is used by all Linux distributions as well as many thousands of corporations worldwide. He works for Google, Inc. who fund him to work full-time on improving Samba and solving the problems of Windows and... Read More →


Thursday March 12, 2015 11:20am - 12:10pm
Studio 3

11:20am

Challenges Implementing a Generic Backup-Restore API For Linux - Matthias G. Eckermann, SUSE
Backup-Restore feels like the unloved step-child of IT: neither cool nor wanted, everybody needs and suffers from implementing it.

For a long time it was sufficient to implement three approaches: reading the filesystem, working with LVs or approaching the block layer.

Today, xfs provides freezing and exporting, btrfs and LVM do CoW and snapshots, btrfs implements send-receive (a potential hook for backup and restore?). Different for scale out filesystems (such as Ceph): are they only backend for backup or do they need to be backed-up as well? How? There is still no generic approach to synchronize active "freezing" of an application with checkpoint-restart on the storage level.

We will discuss some of the existing backup/restore technologies, investigate the differences between the local and distributed cases and explore possibilties of introducing a common API.

Speakers
ME

Matthias Eckermann

Director Product Management, SUSE
Matthias is member of the Product Management team of SUSE. There he is responsible for specification and delivery of SUSE Linux Enterprise Server and related products. Before joining the Product Management team he worked as consultant and senior architect for SUSE on complex and strategic new projects in EMEA, with partners and customers in enterprise and public sector. Matthias has 20 years of experience working with Linux and other Open... Read More →


Thursday March 12, 2015 11:20am - 12:10pm
Loft 2

11:20am

GlusterFS: Data Compliance Infrastructure - Venky Shankar, Red Hat
With recent development efforts to add various "data compliance" mechanisms (such as Data Tiering, BitRot detection, dedup, etc.) to GlusterFS, a common scalable, pluggable & crash consistent filesystem notification infrastructure forms the basis for such techniques. In this session, Venky Shankar explains how Gluster uses it's journaling infrastructure to provide "pure" callback based notification facility via a set of easy to use APIs available as a part of shared library.

Log based synchronous & asynchronous (lazy) replication techniques in GlusterFS have already benefited in many aspects by relying on filesystem journals for data synchronization. The core infrastructure is extended by adding support for data classification (hot/cold data), multi-consumer model, improved performance, out-of-band notifications, etc., yet keeping the design simple and elegant.

This presentation would help users/developers to understand the low level details of the notification infrastructure of GlusterFS and help them understand, use and contribute to it's upstream development.

Speakers
avatar for Venky Shankar

Venky Shankar

Developer, Red Hat
I work as a GlusterFS developer at Red Hat involved mostly in the replication subsystem. I am the lead developer of GlusterFS Geo-replication (asynchronous replication) which was one of the initial features that used journals (log) to synchronize data over high latency links (typically over a WAN), followed by log based synchronous replication module. Both of the above replication strategies were presented (by myself and other developers) in... Read More →


Thursday March 12, 2015 11:20am - 12:10pm
Carver 1

12:10pm

Lunch
Thursday March 12, 2015 12:10pm - 1:40pm

1:40pm

Cache Tiers and Erasure Coding in Ceph - Samuel Just, Red Hat
Ceph is designed around the assumption that all components of the system (disks, hosts, networks) can fail, and has traditionally leveraged replication to provide data durability and reliability. The CRUSH placement algorithm is used to allow failure domains to be defined across hosts, racks, rows, or datacenters, depending on the deployment scale and requirements. Ceph has introduced a storage tiering infrastructure and cache pools that allow alternate hardware backends (like high-end flash) to be leveraged for active data sets while cold data is transparently migrated to slower backends. The combination of these two features enables a surprisingly broad range of new applications and deployment configurations. This talk will cover a few Ceph fundamentals, discuss the new tiering and erasure coding features, and then discuss a variety of ways that the new capabilities can be leveraged.

Speakers
SJ

Samuel Just

Red Hat
I work at Red Hat as the core team PTL for the ceph project. My work focuses on scalability and performance in the ceph rados layer. I spoke at Vault 2015.


Thursday March 12, 2015 1:40pm - 2:30pm
Loft 2

1:40pm

Designing SMR Solutions with libzbc and Open Source Software - Damien Le Moal, HGST
Shingled Magnetic Recording (SMR) is the next generation storage technology for continued improvement in HDD areal density, and offers new opportunities for low cost scale-out storage solutions. The introduction of new standards for managing SMR disks can however impact host software design. New development approaches are required in order to realize the full potential of SMR technology. This speaking session presents the INCITS standards related to SMR and discusses different application design techniques for integrating SMR HDDs into storage solutions.

Speakers
DL

Damien Le Moal

Damien Le Moal is the manager of the System Software Group at HGST Research. He received his PhD in Informatics from Kyoto University in 2010 and has over 15 years of experience in the area of system software for storage solutions. He specializes in operating systems and low-level system software research and development. He is a co-designer and developer of the libzbc open source project and is also involved in research activities including file... Read More →


Thursday March 12, 2015 1:40pm - 2:30pm
Carver 2

1:40pm

Lazytime: Optimizing File Systems By Lazily Writing Back Timestamps - Ted Ts'o, Google
This presentation describes a new way to optimize Linux file system writes by suppressing unnecessary updates to the inode's timestamps. This is done by adding an additional level of inode dirtiness. Currently we have two levels to indicate whether the inode should be written back via fdatasync(2) or not. Lazytime adds a third, weaker, level of dirtiness which suppresses write backs except via an explicit fsync(2) or the entire file system is synced or unmounted. Lazytime was originally conceived as an ext4-specific optimization, since this optimization is especially effective for ext4, but it is now an optimization that can be used by any file system.

Speakers
TT

Theodore Ts'o

Staff Engineer, Google
Theodore Ts'o is the first North American Linux Kernel Developer, and started working with Linux in September, 1991. He previously served as CTO for the Linux Foundation, and is currently employed at Google. Theodore is a Debian Developer, and is the maintainer of the ext4 file system in the Linux kernel. He is the maintainer and original author of the e2fsprogs userspace utilities for the ext2, ext3, and ext4 file systems.


Thursday March 12, 2015 1:40pm - 2:30pm
Studio 3

1:40pm

Providing Atomic Sector Updates in Software for Persistent Memory - Vishal Verma, Intel
Persistent or non volatile memory technologies are often heralded as the next big leap in storage technologies. A lot of existing software can reap the benefits of these technologies if the storage were presented as a standard block device. Traditionally, software expects that sector writes will never be torn - either a sector makes it to storage in its entirety, or not at all. Since persistent memory is inherently byte addressable, simply presenting a persistent memory range as a block device using drivers like 'brd' does not provide us with this atomicity guarantee. In this presentation, Vishal Verma will present a mechanism to solve the sector tearing problem in software using a lightweight system called the Block Translation Table (BTT), which uses existing software constructs to solve some of the unique problems of providing atomic update semantics for byte addressable memory.

Speakers
avatar for Vishal Verma

Vishal Verma

Software Engineer, Intel
Vishal Verma is a developer at Intel's Open Source Technology Center. His previous work includes authoring the SCSI emulation layer for the NVMe driver.


Thursday March 12, 2015 1:40pm - 2:30pm
Carver 1

2:40pm

Improving The Latency of Fsync System Call at EXT4 - Daejun Park, Sungkyunkwan University
Fsync system call is used for data durability in many applications. However, the latency-sensitive synchronous operations can be delayed under ordered mode in ext4. Since the transaction includes irrelevant data and metadata as well as those of fsynced file, the latency of fsync call can be unexpectedly long. We propose a new journaling scheme which records only the fsync related metadata updates without committing unrelated metadata. By removing or mitigating significantly all adversely affecting factors on fsync latency, the proposed technique can reduce the journal write traffic as well as can reduce the fsync latency.

Speakers
avatar for Daejun Park

Daejun Park

PhD Student, Sungkyunkwan University
Daejun Park is a graduate student at Sungkyunkwan University, Korea. His current research area is file systems and flash storage. Currently, he is interested at improving the latency of fsync operation at EXT4 file system.


Thursday March 12, 2015 2:40pm - 3:30pm
Studio 3

2:40pm

Multipathed PCI-e Storage - Keith Busch, Intel
NVM-Express is a standard for PCI-e attached storage, and is fast becoming mainstream with many products available from various vendors today. The next generation of products under development brings new sets of enterprise features that require some enhancements to the software to fully take advantage of its capabilities, one of which is multipathing.

Multipathing PCI-e storage can have interesting applications, especially with larger NUMA configurations. While Linux has much of the components in place already, there are some kernel and driver enhancements required to make Linux readily work with such an NVMe device. This presentation will cover these proposed enhancements from nvme driver, block-multiqueue, and device-mapper's dm-mpath, as well as some user-space tips to get the most out of your h/w, and hopefully get some feedback discussion on the proposals.

Speakers
avatar for Keith Busch

Keith Busch

Software Engineer, Intel Corporation
Keith Busch is a software engineer working for Intel's Non-Volatile Solutions Group (NSG). He develops NVM products, writes the linux drivers and software, and represents Intel at committees and conferences. Keith has written contributions included in NVM-Express specifications, and presented this technology and Linux software supporting it at the Flash Memory Summit (2013 and 2014), LinuxCon Japan 2014, and Linux Vault 2015.


Thursday March 12, 2015 2:40pm - 3:30pm
Loft 2

2:40pm

Pblcache: A Client-Side Persistent Block Cache for the Data Center - Luis Pabon, Red Hat
Cloud data centers deploy a large number of compute nodes to meet increasing application needs. The computes nodes introduce storage cluster contention, resulting in increased application response time. This presentation discusses the design and architecture of Pblcache, a persistent, write-through, block cache for QEMU based on NetApp’s Mercury paper. Pblcache is designed to reduce application response time by providing a shared block cache local to the compute node.

Speakers
avatar for Luis Pabon

Luis Pabon

Principal Software Engineer, Red Hat
Luis Pabon is a Principal Software Engineer member of Red Hat Storage working on OpenStack Storage. Prior to joining Red Hat in February of 2013 he worked at NetApp Advanced Technology Group and at EMC on various storage products.


Thursday March 12, 2015 2:40pm - 3:30pm
Carver 1

2:40pm

XFS: There and Back And There Again - Dave Chinner, Red Hat
The first XFS code was committed to the Irix source repository more than 20 years ago and it has been under constant develpment ever since. First aimed at breaking the 32 bit filesystem size barrier, XFS was designed for lots and large. Today, XFS is widely deployed on Linux and is still the only real choice for people who need to store lots and/or large.

This presentation will walk through some of the significant developments and major deployments during XFS's 20 year history. We will then explore the current state of development, before finally doing some crystal ball gazing to see if XFS is still going to be around in another 20 years time.

Speakers
DC

Dave Chinner

Dave Chinner has an advanced case of Filesystem Developer Syndrome (FDS). Symptoms first developed back in 2002 when employed by SGI to work on NFS. After a confusing diagnosis including NFS, gigabit ethernet driver and TCP/IP stack hacking on Irix, he was discovered in a delirious state, not knowing where he was or what he was doing. Editors with open XFS source code files were found on his workstation, and the diagnosis was clear: it was... Read More →


Thursday March 12, 2015 2:40pm - 3:30pm
Carver 2

3:30pm

Break
Thursday March 12, 2015 3:30pm - 4:00pm
Foyer

4:00pm

Will Your Grandchildren Know What Your Parents Looked Like? - Brian Pawlowski, Senior Vice President and Member of Technical Staff at NetApp
The pace of change in both software and hardware storage technology has accelerated presenting opportunities for saving costs while allowing us to store our increasingly all digital-only information world.  Ensuring data durability and more importantly accessibility remain the biggest long term challenges to business and our culture. How can we tackle this?

Speakers
avatar for Brian Pawlowski

Brian Pawlowski

Senior Vice President and former Chief Technology Officer at NetApp, NetApp
Brian Pawlowski is Senior Vice President and former Chief Technology Officer at NetApp. Pawlowski is currently leading an internal start-up that recently shipped NetApp’s first purpose built All Flash Array Storage product. Since joining NetApp in 1994, he has been involved in the design of high performance, highly reliable storage systems. Pawlowski has been working on open protocols for storage since the 1980’s, was... Read More →


Thursday March 12, 2015 4:00pm - 4:20pm
Space 57

4:20pm

SMR – The Next Opportunity in Storage - Jorge Campello, Global Director of Systems and Solutions, HGST
Shingled Magnetic Recording (SMR) is the next generation storage technology for continued improvement in HDD areal density, and offers new opportunities for storage solutions. In massive, scale-out cold storage applications such as active archive, social media and long-term data storage, SMR HDD-based solutions offer the highest density, lowest TCO and leading $/TB.  This speaking session will clearly articulate the difference in SMR drive architectures and performance characteristics, and will illustrate how the open source community has a distinct advantage in integrating a host-managed platform that leverages SMR HDDs.

Speakers
avatar for Jorge Campello

Jorge Campello

Global Director of Systems and Solutions, HGST
Jorge Campello is the Global Director of Systems and Solutions at HGST Research. He received his PhD in Electrical Engineering from Stanford University in 1999 and has been working in the storage industry over the last 16 years. An SMR evangelist, he was deeply involved in the formation of the standardization efforts around SMR and now directs his efforts at the libzbc open source project. He is also Co-Chair of the Storage Work Group within... Read More →


Thursday March 12, 2015 4:20pm - 4:40pm
Space 57

4:40pm

Open Source vs. Open Standards - Sage Weil, Ceph Principal Architect at Red Hat
Sage Weil, Ceph Principal Architect at Red Hat will discuss open source versus open standards and why open source software will be the key driver in the next revolution of storage.

Speakers
avatar for Sage Weil

Sage Weil

Ceph Principal Architect, Red Hat
Sage Weil originally helped design and build Ceph as part of his graduate research at UC Santa Cruz. Since then, he has continued to lead the engineering effort, building Ceph into an open source, unified storage platform with an emphasis on scalability, reliability, and performance for cloud workloads. He currently works as Red Hat in the Office of Technology, where his primary responsibility is to lead the Ceph development community while... Read More →


Thursday March 12, 2015 4:40pm - 5:00pm
Space 57

5:00pm

How Facebook is Leveraging Linux in the Data Center - Chris Mason, Linux Kernel Engineer at Facebook
Join Linux Kernel Engineer Chris Mason as he discusses how Facebook is leveraging Linux - and a vast open source portfolio - in the data center to connect 1.35 billion people across the globe.

Speakers
CM

Chris Mason

Software Engineer, Facebook
Chris is a Software Engineer on the kernel team at Facebook, and the maintainer of the Btrfs filesystem. He has been working full time on the kernel for over 16 years, and lives in Rochester New York.


Thursday March 12, 2015 5:00pm - 5:20pm
Space 57

6:00pm

Closing Reception
All attendees are invited to the Vault Closing Reception Thursday evening from 6pm-9pm at Whiskey Saigon.  We will meet in the hotel lobby at 5:45pm and walk to the venue together.  Dinner, drinks, and entertainment will be provided. 

Thursday March 12, 2015 6:00pm - 9:00pm
Whiskey Saigon