Loading…
LinuxCon + CloudOpen Japan 2015 has ended

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

Experience Level: Intermediate [clear filter]
Wednesday, June 3
 

11:30 JST

systemd and Containers - Lennart Poettering, Red Hat
systemd is the system and service manager of most modern Linux distributions. Recently the project has been busy with adding first-class support for various container technologies to systemd natively, taking inspiration from Solaris Zones and other container systems. In this talk I want to shed some light on how to use this functionality, and show how the container integration is now available all through the systemd stack. We will particularly closely look at systemd's own minimal container manager "systemd-nspawn", as well as the "machinectl" tool for interfacing with container managers. With recent systemd containers are not a foreign technology running on top of an OS anymore, but are an essential part of the OS itself. This talk will explain the why, and show you how to make the best of it, for your workloads.

Speakers
LP

Lennart Poettering

Principal Software Engineer, Red Hat
Lennart works on systemd, for Red Hat.


Wednesday June 3, 2015 11:30 - 12:20 JST
Jupiter

11:30 JST

Open Source and Samsung: Myth Busting - Ibrahim Haddad, Samsung
True to the style of "MythBusters" Discovery Channel television show, Haddad takes the audience on a journey of Open Source myths busting as it relates to Samsung and Open Source. Instead of crashing cars, racing a plane with a tractor, and testing explosives, we will go over fun stories and examine what it takes to get a company as large as Samsung to adopt open source software for driving innovation and of course covering some myths related to the contribution side, and everything in between.

Speakers
avatar for Ibrahim Haddad

Ibrahim Haddad

Executive Director, LF AI Foundation
Dr. Ibrahim Haddad is a technologist, strategist and an aspiring writer. His focus is on intersections between emerging technology, open source methodology and innovation. He is Vice President of Strategic Programs at the Linux Foundation and the Executive Director of the LF AI Foundation... Read More →


Wednesday June 3, 2015 11:30 - 12:20 JST
Neptune

11:30 JST

Pushing the Limits of Linux on ARM: STM32 and Further SoC Support - Andreas Färber, SUSE
In recent years there has been an explosion of ARM SoCs and matching development boards. In this presentation, Andreas Färber will discuss why it is important for distros such as openSUSE to have board support in the official or "mainline" Linux kernel. Looking beyond well-known ARMv7-A boards such as the Parallella, minimum requirements for Linux on ARMv7-M will be discussed and projects presented for mainline Linux on STM32 and further microcontrollers.

Speakers
avatar for Andreas Färber

Andreas Färber

Project Manager arm64, SUSE
Andreas has been with SUSE since 2011, working on KVM virtualization as an engineer and regularly speaking at KVM Forum. He has been behind the openSUSE arm port since its restart in 2011, making Linux and openSUSE run on various boards and devices. Since 2017 he is the Project Manager... Read More →


Wednesday June 3, 2015 11:30 - 12:20 JST
Mars

11:30 JST

SL[AUO]B: Kernel Memory Allocators for Smaller Objects - Christoph Lameter
An introduction to the design and ideas behind the 3 small object memory allocators in the Linux kernel. SLAB is the original memory allocator based on ideas also used for Solaris. SLOB is a based K&R allocator originating even further in the past. SLUB is an allocator written for minimal queuing focusing on low instruction overhead in the fast paths instead.

Slab allocators are providing basic memory allocation to kernel code. Often these allocations may occur in performance sensitive code (especially in the network and storage layer). Kernel performance is significantly affected by the slab allocators architecture and features. The talk provides an overview of the slab allocator services available in the Linux kernel and covers the most frequent use cases. Also includes a discussion of current issues and a roadmap of development in the past and into the future.

Speakers
avatar for Christoph Lameter

Christoph Lameter

R&D Team Lead, Jump Trading LLC
Christoph Lameter is working as a lead in research and development for Jump Trading LLC (an algorithmic trading company) in Chicago and maintains the slab allocators and the per cpu subsystems in the Linux Kernel. He contributed to a number of Linux projects since the initial kernel... Read More →


Wednesday June 3, 2015 11:30 - 12:20 JST
Orchid 1 & 2

14:00 JST

How To Develop Magnum(Container as a Service) - Motohiro Otsuka, NEC & Madhuri Kumari, NEC
Magnum is a new API Service under development by the OpenStack Containers Team
to make container management tools such as Docker and Kubernetes available as a
first class resource in OpenStack. Start from "What is Magnum?" and then going to includes how to use Magnum, how to develop (on your Linux/MacBook) and how to contribute.

Magnum uses Heat,Ironic (via Nova) and of course Docker and Kubernetes, so building development environment is a little difficult. Using Vagrant and Devstack is easy way to build development environment, because you can build a environment all over again easily. After this session, you will be able to contribute Magnum easily.

Speakers
MK

Madhuri Kumari

Madhuri is a Software Engineer at NEC Technologies India Pvt Ltd. She has experience in the storage and cloud domain. She is an active contributor in Openstack Projects. She has also contributed to Magnum, Swift, Nova.She is a core in Openstack Container Project "Magnum".
MO

Motohiro Otsuka

Motohiro is developing OpenStack in the community, and he is a core developer of OpenStack Magnumproject project.


Wednesday June 3, 2015 14:00 - 14:50 JST
Jupiter

14:00 JST

Dynamic Tracing Updates - Masami Hiramatsu, Hitachi
Perf-probe dynamic event tracing is a fundamental feature for behavior analysis, performance analysis, and debugging. Recently, it has many updates and planned works as ARM porting/enhancement, perf-probe caching (debuginfo caching), and SDT support(for uprobe-events) etc. This presentation will show you these efforts including ongoing works.

Speakers
MH

Masami Hiramatsu

Researcher, Hitachi Ltd.
Masami Hiramatsu is a Japanese kernel maintainer of kprobes/ftrace/perf-probe etc. He is working for Hitachi Ltd. and is a researcher in Yokohama Research Laboratory. He started working on Linux kernel with Kernel Tracing (LKST) at 2002, and joined to SystemTap development and became... Read More →


Wednesday June 3, 2015 14:00 - 14:50 JST
Mars

14:00 JST

What Is Suspend-To-Idle and How to Make It Work - Rafael J. Wysocki, Intel
Generally, suspend-to-idle is a system suspend variant whose last stage is to leave processors in idle states instead of taking them offline (except for one) and calling platform-specific code or firmware to turn off whatever is still drawing power. Although the idea is rather straightforward, implementing it in practice turned out to be quite challenging. In paticular, it involved reworking system wakeup interrupts handling in the IRQ subsystem and adding support for quiescing timers and suspending timekeeping from within the kernel's idle loop. Doing that work allowed developers to extend their understanding of some issues related to system suspend and will hopefully result in better support for that feature overall going forward. I will discuss the work on implementing support for efficient suspend-to-idle that has been in progress since the early 2013 and the lessons learned from it.

Speakers
avatar for Rafael J. Wysocki

Rafael J. Wysocki

Software Engineer, Intel OTC
Rafael maintains the Linux kernel's core ACPI and power management code, including the core infrastructure for IO device PM, CPU PM and system suspend/hibernation. He works at Intel Open Source Technology Center as a Software Engineer focusing on the mainline Linux kernel. Rafael... Read More →


Wednesday June 3, 2015 14:00 - 14:50 JST
Soleil

14:00 JST

Kernel Security Hacking for the Internet of Things - Daniel Sangorrin, Toshiba
The so-called 'Internet of Things' will connect embedded computing devices of all kinds, allowing for the development of powerful services that were not possible before. On the other hand, the interconnection of traditionally isolated devices will pose new security challenges that require consideration. In this tutorial, Daniel will focus on showing you how to extend the Linux kernel with functionality to detect attacks against connected embedded computing devices. The purpose of his talk is to raise concern about embedded security; and to foster the development and sharing of security technology able to protect the valuable assets of our society.

Speakers
avatar for Daniel Sangorrin

Daniel Sangorrin

Expert, Toshiba corp.
Daniel Sangorrin works for Toshiba corp. as an operating systems researcher with a focus on real-time embedded systems. He received a Ph.D degree in computer science from Nagoya University, and has been a speaker in several international conferences and open source events.


Wednesday June 3, 2015 14:00 - 15:50 JST
Orchid 1 & 2

15:00 JST

Enhanced Live Migration for Intensive Memory Loads in the Cloud - Mario Smarduch, Samsung
Live Migration is essential to deliver a seamless cloud user experience for both load balancing and rapid scaling. However, aggressive memory workloads can lead to long pauses, resulting in service outages, or require the Guest to significantly lower the CPU load.

This presentation covers the Linux Kernel KVM 'rapid migration' support implemented for ARM v7 and v8. Several new features enable migration of memory intensive Guests. This presentation will dive deep into the subtleties of live migration, Linux kernel KVM ARM32 and ARM64 MMU, dirty page logging/live migration, and QEMU support. In addition a newly designed test environment to validate live migration and eliminate latent faults will be presented

If you manage memory intensive guests and want to gain a solid understanding of Live Migration and the new ARM64 Linux Kernel KVM architecture, you must attend this talk.

Speakers
MS

Mario Smarduch

Senior Virtualization Architect, Samsung
Mario Smarduch is a Senior Virtualization Architect at Samsung's Open Source Group. Currently, he's working on ARM-KVM features and optimizations for Samsung products. He's also engaged in the Linux kernel KVM open source community, contributing some key features to KVM ARMv8 and... Read More →


Wednesday June 3, 2015 15:00 - 15:50 JST
Jupiter

15:00 JST

NFV Infrastructure Manager with High Performance Software Switch Lagopus - Hirofumi Ichihara, NTT
Network performance is important requirement to achieve Network Function Virtualization (NFV). Open vSwitch and Linux bridge are able to be selected as OSS virtual switch on hypervisor in OpenStack taken notice as the NFV infrastructure manager. However, it's difficult that these software switches fulfill network performance of NFV requirement. Some techniques are proposed to improve the network performance. DPDK is one of the techniques. This presentation will give an overview of high performance software switch Lagopus with DPDK, an introduction how to control Lagopus with an infrastructure manager and integration OpenStack.

Speakers
HI

Hirofumi Ichihara

Software Engineer, NTT corporation


Wednesday June 3, 2015 15:00 - 15:50 JST
Orchid 3

15:00 JST

Visualizing Used Symbols for Linux Kernel Modules for License Compliance - Armjin Hemel, Tjaldur Software Solutions
The Linux kernel has a mechanism to expose certain symbols only to modules that declare themselves as licensed under GPL. If they are not, then these symbols cannot be used. For many people it is unclear how to find this out. In this talk I want to present a method that makes relationships between Linux kernel modules, the declared licenses and the licenses of the symbols more clear by visualizing them.

Speakers
AH

Armijn Hemel, Tjaldur Software Governance Solutions

Tjaldur Software Governance Solutions
Armijn Hemel, MSc, is a Dutch technologist, specialising in license compliance engineering and supply chain management. As a former member of the core team of gpl-violations.org he has intimate knowledge of license enforcement, common mistakes in supply chains and resolution of these... Read More →


Wednesday June 3, 2015 15:00 - 15:50 JST
Soleil

16:20 JST

Solving the Package Problem - Joe Brockmeier, Red Hat
 In the beginning there was RPM (and Debian packages) and it was good. Certainly, Linux packaging has solved many problems and pain points for system admins and developers over the years -- but as software development and deployment have evolved, new pain points have cropped up that have not been solved by traditional packaging.

In this talk, Joe Brockmeier will run through some of the problems that admins and developers have run into, and some of the solutions that organizations should be looking at to solve their issues with developing and deploying software. This includes Software Collections, Docker and Rocket containers, OStree and rpm-ostree, Platform-as-a-Service, and more. 

Speakers
avatar for Joe Brockmeier

Joe Brockmeier

Red Hat, Red Hat
Joe Brockmeier is a long-time participant in open source projects and former technology journalist. Brockmeier has worked as the openSUSE Community Manager, is an Apache Software Foundation (ASF) member, and participates heavily in the Fedora Cloud Working Group. Brockmeier works... Read More →


Wednesday June 3, 2015 16:20 - 17:10 JST
Jupiter

16:20 JST

cgroup Support for Writeback - Tejun Heo, Facebook
cgroup's support for IO resource provision was missing a major piece - the block layer couldn't tell to which cgroup a given writeback IO should be attributed. All non-direct write IOs generated through write(2) and dirtied mmap pages were charged against the root cgroup defeating the purpose of IO provisioning. cgroup is finally growing writeback support which will not only improve IO provisioning but also make memory provisioning behave correctly with respect to the IO pressure in the cgroup. This presentation examines what was broken and how the new cgroup writeback support closes the gap.

Speakers
TH

Tejun Heo

Software Engineer, Facebook
Tejun has been working on various aspects of Linux kernel since 2005 and is currently maintaining percpu memory allocator, control groups, and workqueue. He currently works as a software engineer for Red Hat.


Wednesday June 3, 2015 16:20 - 17:10 JST
Neptune

16:20 JST

RAS Enhancement Activity Updates for Mission-critical Linux Systems - Hidehiro Kawai, Hitachi
Since 2013, Hidehiro Kawai et al. has tried to push RAS features and bug fixes, originally developed for their mission-critical server for control systems, into the mainline kernel. In this presentation, he will talk about what sort of features are required for those servers, what have been done in the mainline, and what would be done in the next. He would like to share remaining issues to promote the activity. This is positioned as an update of Yoshihiro Yunomae's presentation in LinuxCon JP/NA in the last year.

Speakers
HK

Hidehiro Kawai

Researcher, Hitachi, Ltd.
Hidehiro Kawai has been working at Hitachi since 2004 and developing Linux kernel and related tools. Currently, he is working for Linux enhancement in the field of control systems.


Wednesday June 3, 2015 16:20 - 17:10 JST
Orchid 1 & 2

17:20 JST

nfs-ganesha and Clustered NAS : Scale Out Your Storage - Anand Subramanian, Red Hat
In this session we will discuss some important aspects in using the extensible user-space NFS server nfs-ganesha to scale out your storage needs. Nfs-ganesha supports NFSv3,v4,v4.1,v4.2 and pNFS. It also has an easily pluggable architecture called FSAL (File System Abstraction Layer) which enables seamless integration with many filesystem backends (GlusterFS, Ceph, GPFS, Lustre, XFS etc.). We will take a detailed look at the newly introduced CMAL (Cluster Manager Abstraction Layer) allows nfs-ganesha to operate in clustered mode and integrate with a variety of Cluster Managers, allowing for a flexible Clustered NAS solution with multi-head NFS configuration (active-active), distributed lock manager and recovery. We will also look at the example of nfs-ganesha and GlusterFS integration as a clustered storage solution and an open source pNFS solution. 

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... Read More →


Wednesday June 3, 2015 17:20 - 18:10 JST
Jupiter

17:20 JST

Accelerated Data Processing on SoC with FPGA - Marek Vašut, DENX Software Engineer
Accelerated data processing used to be dominated by SoCs with GPUs in the embedded segment. Recently, an exciting combination of SoC with FPGA became available from both major FPGA vendors. In this talk, Marek will explain the pros and cons of using the embedded FPGA instead of the GPU for computationaly-heavy tasks which cannot be handled by the CPU. Since communicating with the FPGA and transporting data between the FPGA and memory requires a non-trivial software infrastructure, which can introduce unexpected performance surprises, different approaches to this task will be explained. In the end, Marek will present the possibility of accelerating OpenCL kernels on the FPGA.

Speakers
avatar for Marek Vasut

Marek Vasut

Software engineer, Self employed
I have been a contractor for multiple companies for many years. My primary responsibility is designing and implementing customer-specific functionality. One important aspect of my work is leveraging the benefits of working inside the mainline Linux, U-Boot and OE / Yocto Project... Read More →


Wednesday June 3, 2015 17:20 - 18:10 JST
Orchid 1 & 2

17:20 JST

CRIU in Safety Systems - Sungho Kim, Hitachi
While Checkpoint-Restart is mostly used to realize zero down-time redundancy, maintaining quorum could be another use case for industrial safety systems requiring redundancy and voting feature. Especially, voting needs proper quorum to judge safety in those systems. Therefore, whenever faulty node exist, new node is required to be added to keep proper quorum without stopping the system. In this presentation, Sungho KIM will share the use case of Linux Checkpoint-Restart functionality, CRIU, in application of safety systems and discuss feature enhancement from view points of Real-time. To copy the processing state of running node to new node joining to the quorum, CRIU will provide some solutions. However, copying process tends to impact on real-time of running nodes. Here, possible problems applying CRIU in this case and the way to avoid the problems are to be explained in a detail way.

Speakers
SK

Sungho Kim

Sungho Kim works for Hitachi Ltd, Yokohama Research Laboratory as a software engineer on software platform for control systems with Linux.


Wednesday June 3, 2015 17:20 - 18:10 JST
Soleil

17:20 JST

Finding Stupid Security Bugs - Armijn Hemel, Tjaldur Software Governance Solutions
With more and more devices being connected to networks (Internet, cellular, sensornetworks, etc.) security is becoming a bigger topic. Many of the security bugs in devices like consumer electronics go unnoticed and are not covered by for example CVE reports, or are never even reported. In this talk I want to propose a few methods that help finding security bugs in devices, hopefully before they become a problem.

Speakers
AH

Armijn Hemel, Tjaldur Software Governance Solutions

Tjaldur Software Governance Solutions
Armijn Hemel, MSc, is a Dutch technologist, specialising in license compliance engineering and supply chain management. As a former member of the core team of gpl-violations.org he has intimate knowledge of license enforcement, common mistakes in supply chains and resolution of these... Read More →


Wednesday June 3, 2015 17:20 - 18:10 JST
Neptune

17:20 JST

Power Efficient Idle Injection - Jacob Pan, Intel
While optimizing for performance are common, many use cases demand just the the opposite: injection of idle cycles. The motivations for doing so range from power capping, thermal management, pay-per-use, and real-time throttling. Idle injection mechanisms are also diversified and provide uneven results in terms power efficiency, performance and user experience impact.

This talk focuses on the power awareness aspect of idle injection, helping audience to connect the kernel mechanism with the underlining hardware. Bridge the understanding of idle cycles vs idle states, scheduling workload vs. hardware coordination, and power saving vs. CPU topology.

Speakers
avatar for Jacob Pan

Jacob Pan

Linux Kernel Developer, Intel Corporation
Jacob is a veteran Linux kernel developer at Intel. His most recent interest and work are on Shared Virtual Address/Memory as well as the IOMMU subsystem in general.Prior to that, Jacob contributed to power management, device drivers, interrupt, timers, and X86 core.


Wednesday June 3, 2015 17:20 - 18:10 JST
Mars
 
Thursday, June 4
 

11:30 JST

Development of Container/Docker for Enterprise Use - Gui Jianfeng, Fujitsu
Fujitsu has been working for linux container for several years. We have been working for a feature for pid namespace in this half year and our work in the kernel will finish soon. The feature is to show pid releationship in different namespaces, we consider it's required for enterprise. We're now planning to integrate the feature into docker. In this session, we'll talk about the design and feature/purpose of pid-matching function.

Docker is rapidly growing but it seems contribution by traditional enterprise companies are not very much (at 2015/Feb). We'd like to talk about our contribution plans to docker/linux kernel as a development team of Fujitsu, e.g. resource control, debug, stability.

Speakers

Thursday June 4, 2015 11:30 - 12:20 JST
Jupiter

11:30 JST

Demand for Linux According to the Future Challenges - Tsugikazu Shibata, NEC
Linux is running variety of place. Large trading system such as Tokyo, London and New york Stock Exchange, Android devices used by Billions of people world wide every day, over 97% of Super computer system give us important results. Linux is running many place and giving us number of great benefits by the result of excellent works of Linux kernel development community. Also, Linux is expected to use for the future technologies such as Docker, Cloud computing, IoT, Drone, Automotive and so on. This presentation discuss about challenges for Linux according to such new technologies. that will not just each of technical features but also including challenges for Industries as the Ecosystem viewpoint such as Long Term support, Quality and so on.

Speakers
avatar for Tsugikazu Shibata

Tsugikazu Shibata

Chief Advanced Technologist, NEC
Tsugikazu Shibata is leading LTSI Project. He has been working on coordinating the relationship among the industry, company and community. He is an active member of various and wide range of Open Source Projects from Embedded to Cloud Computing. He has been spoken many of Linux and... Read More →


Thursday June 4, 2015 11:30 - 12:20 JST
Neptune

11:30 JST

Developing and Monitoring Nodejs/Javascript IoT Apps - Tetsuo Seto, Concurix
Nodejs is entering a new era with the announcement of newly formed Nodejs Foundation. The Linux Foundation will advise on structuring and running the organization. Intel released last year its Edison embedded platform built on top of Yocto with Nodejs built-in targeted for Internet of Things, then revealed Curie this year.

In this presentation, Tetsuo Seto will review one of the best techniques (source code available on github) for profiling and monitoring Nodejs applications as well as garnering and analyzing data emitted by Internet-of-Things application prototyped on Intel Edison. In addition to showcasing the powerful combination of Nodejs/Javasscript, Intel Edison and the app monitoring and data analytics solution, Tetsuo's presentation will include a glimpse of IoT demo for Smart Home as well as Enterprise/Office use.

Speakers
TS

Tetsuo Seto

I'm Tetsuo Seto, VP of Analytics at Concurix Corporation specializing in Nodejs application profiling, monitoring and data analytics technology. I've presented the profiling and monitoring solution at NodeFest Tokyo in Nov. 2014. I have a repository on github.com as Setogit. For... Read More →


Thursday June 4, 2015 11:30 - 12:20 JST
Soleil

11:30 JST

Speed-up perf Tools Using Multi-Thread - Namhyung Kim, LG
The linux perf tools is a modern way of profiling system performance with hardware and/or kernel support. However, on a large profiling session, it could take a long time to analyze the data during report time because it processes the data in a single thread. Since most of those analysis can be partitioned for each cpu or thread, it can make use of multi-threading to speed up the processing.

In this talk, Namhyung will look at the internals of perf tools and changes required to support parallel processing during his work.

Speakers
NK

Namhyung Kim

Software Engineer, Google
Namhyung Kim is a software engineer at Google and have been involved in the development of the perf and ftrace since 2012. With this background he started the uftrace project to improve tracing of userspace programs.


Thursday June 4, 2015 11:30 - 12:20 JST
Orchid 1 & 2

14:00 JST

Tales from the OpenStack's Gate: How Debugging the Gate Helps Your Enterprise - Matthew Treinish, OpenStack
OpenStack is incredibly complex with tons of moving parts. As part of the community's integration system, often called the gate, we run tests against a running OpenStack cloud. However, sometimes things fail for reasons unrelated to the patch being tested. These failures, often called "flakey failures" by developers, are normally an exposure of interesting timing problems in OpenStack and they can often be very tricky to debug. However, we've found that almost all of these issues and the difficulty in debugging them also apply to real deployments. This talk will go over the mechanics of debugging gate failures. As part of the discussion comparisons will be drawn to show how similar the process of debugging gate failures is to production OpenStack issues. To reinforce that anything we see in the gate most likely is being dealt with by users and operators.

Speakers
avatar for Matthew Treinish

Matthew Treinish

Software Engineer, IBM Research
Matthew Treinish has been working on and contributing to Open Source software for most of his career. Matthew currently works for IBM Research developing open source software for quantum computing. He is also a long time OpenStack contributor and a former member of the OpenStack TC... Read More →


Thursday June 4, 2015 14:00 - 14:50 JST
Jupiter

14:00 JST

Implementing a Standard API - Mark Carlson, Toshiba
 There are open source projects where the API from the code *is* the standard, but what about de jure standards such as for Cloud Storage? The Cloud Data Management Interface (CDMI) is an ISO/IEC standard that is starting to see wide adoption with over 20 different implementations all interoperating. This talk will overview some of the open source implementations of CDMI, and detail the open source reference implementation available from the Storage Networking Industry Association (SNIA).

The code is written in Java and is available under the BSD license. Over 1000 downloads of this code have been from nearly every country. Sitting on top of any filesystem, the CDMI protocol is supported, placing cloud objects on the underlying filesystem.

Speakers
avatar for Mark Carlson

Mark Carlson

Principal Engineer, CoChair SNIA TC, Toshiba Memory
Co-Chair SNIA Technical Council Co-chair Object Drive TWG Co-chair Cloud Storage TWG Principal Engineer, Industry Standards at Toshiba, has more than 35 years of experience with Networking and Storage development and more than 18 years experience with Java technology. Mark was one... Read More →


Thursday June 4, 2015 14:00 - 14:50 JST
Neptune

14:00 JST

Information Leak Tracing System - Kazuki Omo, OSS Security Comunity
Current main concern about security is "Cyber Attack" and "Information leak". And attacker sold the leaked information to third pary, then other third party use obtained leaking information for their sales. Then it's better to focus admin's mind from "how to protect" to "how to track leaked information path". In this presentation, Kazuki Omo will suggest the new system for contain original information in file, and discuss how to track the information leak path by the system.

Speakers
avatar for Kazuki Omo

Kazuki Omo

Executive Officer, SIOS Technology Inc.
Over 20 years experience in Unix/Linux/Windows system and many of Security related product. Working for OSS community over 15 years. - Published SELinux and related security articles from 2004-2018. - Presentation on Open Source Summit Japan 2017 "OSS CVE Trends". - Presentation on... Read More →


Thursday June 4, 2015 14:00 - 14:50 JST
Soleil

14:00 JST

Solving Device Tree Issues - Frank Rowand, Sony Mobile
Usage of device tree is mandatory for all new ARM systems. But the implementation of device tree has lagged behind the mandate. The first priority has been correct function. Lower priorities include device tree validation and facitilities to debug device tree problems and errors. This talk will focus on the status of debug facilities, how to debug device tree issues, and debug tips and tricks. Suggestions will be provided to driver writers for how implement drivers to ease troubleshooting.

Speakers
avatar for Frank Rowand

Frank Rowand

Sony
Frank has meddled in the internals of several proprietary operating systems, but has been loyal to the Linux kernel since 1999. He has worked in many areas of technology, including performance, networking, platform support, drivers, real-time, and embedded. He is currently employed... Read More →


Thursday June 4, 2015 14:00 - 14:50 JST
Orchid 1 & 2

15:00 JST

Hardware Accelerartion for NFV - Yuhong Tao, Huawei
Tranditional teloc equipments use lots of hardware accelerators, how to leverage these accelerators becomes important in Network Function Virtualization(NFV). We are trying to implement a software solution based on KVM, to offload specific processing like encrypt, compressing/uncompressing, media transcoding etc, this prestation will introduce new virtio devices which can be used as a hardware accelerator offloading device.

Speakers
YT

Yuhong Tao

Yuhong Tao, PHD. a senior software engineer of Huawei. Huawei is a telecommunication technology solution provider, one of the world 500 strong enterprise. Yuhong Tao is working at NFV for this moment, his interest is using hardware ability to accelerate veritualized-network-function... Read More →


Thursday June 4, 2015 15:00 - 15:50 JST
Orchid 1 & 2

16:20 JST

Live Migrate Guests w/PCI Pass-Through Devices - Taku Izumi, Fujitsu
Live migration is one of KVM key features. PCI pass-through is also useful feature for those who require high I/O performance on guests. However thise two features don't go together. Currently we can't live-migrate guests with PCI pass-through devices. We're now addressing
to lift this limitation. In this session, we'd like to illustrate our proposal and current status.

Speakers
TI

Taku Izumi

Fujitsu Limited
I've been working for Fujitsu since 2003. I had worked for support service of Linux network drivers for several years. And now, I've been working for Linux development team in Fujitsu for these several years.


Thursday June 4, 2015 16:20 - 17:10 JST
Orchid 3

16:20 JST

Sheepdog is Ready: Distributed Block Storage is Turning From Experiment to Commercial Use - Teruaki Ishizaki, NTT Labs
Sheepdog enables to make a scalable and reliable storage system with using IA servers. Sheepdog provides many access heads, Qemu, iSCSI, ObjectStorage, Cloud Systems use it as a unified storage. Sheepdog has been developed for improving reliability and stability. So we can use Sheepdog for more critical systems and now is ready for commercial use! In this talk, the recent results of development are presented. Remarkable results are:
  1. iSCSI multipath: Sheepdog has introduced the algorithm for managing state of tgtd's connections. Now iSCSI multipath with active-backup is supported and tgtd isn't a SPOF.
  2. Improved GC implementation: Sheepdog couldn't track complicated snapshot relations, so data objects of deleted volumes sometimes remained falsely. Now disk consumption of deleted volume are dramatically reduced.
In addition, performance tests with SSD are presented.

Speakers

Thursday June 4, 2015 16:20 - 17:10 JST
Jupiter

16:20 JST

Btrfs - Current Status and Future Prospects - Qu Wenruo, Fujitsu
Join us as we discuss the most recent state of the btrfs filesystem, including:
  • What is Btrfs
  • Essential features
  • Why XFS and ext4 with LVM are insufficient
  • The status of Btrfs from the perspective of enterprise usage Functions, stability, performance, and so on
  • What has changed recently
  • Improve RAID5/6 support, fsck, qgroup, and so on
  • How/what we will do next
  • Improve SSD related functions
  • Improve documentation
  • Improve the stability
 

Speakers
QW

Qu Wenruo

Software Engineer of Fujitsu.


Thursday June 4, 2015 16:20 - 17:10 JST
Neptune

16:20 JST

Exciting Developments in Linux Tracing - Elena Zannoni, Oracle
Tracing in Linux today has become quite main stream. Linux has come a long way from when the first attempt to insert tracing infrastructure and tools in the kernel tree was made several years ago. But where are we now? While existing tools are being refined and more sophisticated features are added to them, there are still fundamental areas where improvement is happening and exciting new approaches are surfacing, such as using the BPF (Berkeley Packet Filtering) infrastructure for tracing purposes, for instance. This talk is a new and revised version of my previous tracing talk with specific emphasis on new developments as opposed to providing an overview of the field.


Speakers
avatar for Elena Zannoni

Elena Zannoni

Director of the Linux Tools and Languages Team, Oracle Corporation
Elena Zannoni is the manager for the Linux Toolchain and Tracing team at Oracle. The team covers the GNU toolchain and DTrace for Linux, among other things. Elena was one of the original GDB global maintainers and has spoken worldwide on topics related to tracing at many conferences... Read More →


Thursday June 4, 2015 16:20 - 17:10 JST
Soleil

16:20 JST

Hints and Tips to Bring your Raspberry Pi Project Into "Real World" Operation - Masahiro Furutera
If you consider your DIY project into 'real world' operation, there are a lot of things to be considered and need to jump several varriers. Some are similar to commertial product development, but others are different. The speaker provides Hint and Tips according to his experiences for Raspberry Pi/Raspbian (Debian Wheezy) based 'Gate Door' contoroller. This presentation covers folloing topics;
  • Law regulations (mostly in Japan)
  • Safety consideration
  • Produt Liability consideration
  • Fale Safe design and planning for out-of-service, how to improve availability
  • Life cycle and repairing consideration
  • How to simplify and freeze your application as an embeded system
  • Remote Operation consideration
The speaker also shows his test system and demonstrates remote operation capability.

Speakers
avatar for Masahiro Furutera

Masahiro Furutera

CMO, Time Machine Inc.
35+ years experience for IT industry from development to marketing. 2nd level pre sales technical support and marketing for IBM Power and AIX(IBM Unix).Working with Linux Community as a Linux Allians manager for last 5 years at IBM.After retiring IBM, working for personal busuiness... Read More →


Thursday June 4, 2015 16:20 - 17:10 JST
Orchid 1 & 2

16:20 JST

Successful Engagement with Open Source Communities - Tim Bird, Sony Mobile
The CE Workgroup of the Linux Foundation has initiated a project to overcome obstacles, within the embedded industry, to mainlining source code for the Linux kernel by corporate developers. In this presentation, Tim Bird will describe the activities of that project, including our "obstacles to mainlining" survey and white paper, online resources for aiding corporate developers in mainlining, and Tim's analysis of multiple Linux source trees from mobile phone vendors. Included in the talk will be a list of lessons learned and best practices for community interaction, that Tim has compiled over the years.

Speakers
avatar for Tim Bird

Tim Bird

Senior Software Engeineer, Sony Mobile
Tim Bird is a Senior Staff Software Engineer for Sony Corporation, where he helps Sony improve the Linux kernel for use in Sony's products. Tim is also the Chair of the Architecture Group of the CE Working Group of the Linux Foundation. Tim has been working with Linux for over 20... Read More →


Thursday June 4, 2015 16:20 - 17:10 JST
Mars

17:20 JST

FloatingIP Enhancement for Public Cloud Infrastructure - Yushiro Furukawa, Fujitsu
In a public cloud environment, floating IP is used for VM's IP address. If a VM needs the Internet access, global IPs are used as floating IP. In the situation, however, there's something to improve with the current floating IP mechanism.
  1. In a typical public cloud environment, there're some network hardware that are connected to the external network. With the current floating IP mechanism, global IP is required for each hardware, which can be considerable waste.
  2. VM instances that have a floating IP can access the network hardware, which may be a concern for the cloud provider.
In this presentation, we propose an enhancement for Neutron as a solution for both of them.

Speakers
avatar for Yushiro Furukawa

Yushiro Furukawa

Software Engineer, Fujitsu
"Yushiro Furukawa has been working for Fujitsu for 8 years. He worked on OpenStack neutron/fwaas/ironic for 4 years. Currnntly, he's focusing on Kubernates.


Thursday June 4, 2015 17:20 - 18:10 JST
Jupiter

17:20 JST

An Introduction to Greybus - Greg Kroah-Hartman, Linux Foundation
Greybus is the name for a new application layer protocol on top of
Unipro that controls the Ara Phone from Google.  This protocol turns a
phone into a modular device, allowing any part of the system to be
hotplugged while the phone is running.  This talk will describe what
this protocol is, why it was designed, and give the basics for how it
works.  It will discuss how this is implemented in the Linux kernel, and
how it easily bridges existing hardware like USB, I2C, GPIO and others
with little to no changes needed to existing kernel drivers.

Speakers
avatar for Greg Kroah-Hartman

Greg Kroah-Hartman

TAB Member, Linux Foundation
Greg Kroah-Hartman is among a distinguished group of software developers who maintain Linux at the kernel level. In his role as a Linux Foundation Fellow, he continues his work as the maintainer for the Linux stable kernel branch and a variety of subsystems while working in a fully... Read More →


Thursday June 4, 2015 17:20 - 18:10 JST
Mars

17:20 JST

Introduction to GPUs and the Free Software Graphics Stack - Alexandre Courbot, NVIDIA
The GPU (Graphics Processing Unit) is one of the most complex and interesting components of a modern personal computer, be it a desktop machine or a mobile SoC. This talk is intended to be a gentle introduction to GPU history, how modern GPUs work, and how the Linux graphics stack is built to take advantage of all their features. Due to obvious time limits we will only scratch the surface, but we hope to lift some of the mystery that shrouds GPUs and encourage curious but potentially scared developers to study and contribute to the open-source graphics stack.

The talk will cover how 3D graphics are rendered, how GPUs accelerate the process, and the basic layout of the kernel and user-space drivers. NVIDIA GPUs and the Nouveau Free driver will be used as references for this talk, but the general principles should be applicable to any modern GPU.

Speakers
avatar for Alexandre Courbot

Alexandre Courbot

Software Engineer, NVIDIA
Alex is employed by NVIDIA to support the latest Tegra GPUs with the Nouveau free driver. His other kernel-related tasks include co-maintaining the Tegra architecture and GPIO subsystems.


Thursday June 4, 2015 17:20 - 18:10 JST
Orchid 1 & 2
 
Friday, June 5
 

11:30 JST

The OpenDOF Project as an ECHONET Bridge for Smart Homes - Bryant Eastham, Panasonic R&D
The OpenDOF Project provides a secure, open-source framework for easily building distributed, interoperable systems that include devices, gateways, and cloud services for the Internet of Things. With a consistent, flexible object model these systems include data collection as well as remote control.

One of the key capabilities of the framework is the creation of bridges between different transports and application protocols. The benefit of these bridges is compatibility with other OpenDOF-based cloud services, as well as its security and object model.

This presentation will introduce the OpenDOF Project as well as discuss an existing ECHONET bridge and the technical issues that needed to be resolved in its development. These types of bridges are key to the development of Smart Homes.

Friday June 5, 2015 11:30 - 12:20 JST
Mars

14:00 JST

Converting Existing Installed OS and Applications to Docker - Marc Merlin, Google
Google has been using the same shared server image between the booting OS and the applications that runs in containers. Container solutions like docker were designed to let you run an application on a vanilla OS image. This is great if you want to totally decouple a new application from the underlying OS. This is not great if you don't want to have to maintain 2 operating systems (the host one and the container one) if you already have well customized OS, and if you simply want to run some existing applications in a container. This gives you the security benefit from resource/disk/network separation between untrusted applications and the rest of your operating system without having to start from scratch with a base OS and lots duplicate packages between the container and the host OS. More importantly this lets you switch existing working applications to a state where half their functionality (like the web facing portion with untrusted php code) runs in a read only container, while the backend that needs access to local disk and network resources, run outside the container, and post the data to a filesystem that is only read only in the container but sufficient for serving results.

You also benefit from completely controlling the container image and not relying on an image retrieved from the internet that may not be trustworthy by reusing the exact same OS image inside and outside the container, also minimizing upgrade time and cost.

My talk will show how to use docker to achieve those goals and get a chance to containerize applications that you probably wouldn't get around to run in containers otherwise

Speakers
avatar for Marc Merlin

Marc Merlin

Linux Geek / BOFH, Google
Marc has been using Linux since 0.99pl15f (slackware 1.1.2, 1994), both as a sysadmin and userland contributor. He has worked for various tech companies in the Silicon Valley, including Network Appliance, SGI, VA Linux, Sourceforge.net, and now Google. He has been giving Linux related... Read More →


Friday June 5, 2015 14:00 - 14:50 JST
Orchid 1 & 2

14:00 JST

Git Internals - Graphs, Hashes, and Compression - John Britton, GitHub
Git is a version control system built on well known patterns in computer science. John will take you to the very bottom. The floor. The code. The algorithms. The directed acyclic graph of hashed bit sequences made efficient through LZW compression and deferred garbage collection determined by node reachability via hash relationships. We want and need to know precisely how things work. Let’s dig into the guts of Git.

Speakers
JB

John Britton

John Britton is a developer and community builder, active in both open source and open education. As Education Liaison at GitHub he's working to improve computer science education by bringing the principles of open source into the classroom.


Friday June 5, 2015 14:00 - 14:50 JST
Neptune

14:00 JST

Using Rust to Build the Next Generation Web Browser - Lars Bergstrom, Mozilla & Mike Blumenkrantz, Samsung
This talk will discuss how Rust is being used to develop Servo, a web browser optimized to run on both desktop and embedded devices. Rust is an alternative to C for systems programming which provides memory safety guarantees while maintaining a similar level of performance. Rust is being leveraged in Servo to parallelize operations wherever possible. In this case, parallelization results in greatly reduced power consumption, making Servo a key player in the mobile browser space.

Speakers
avatar for Lars Bergstrom

Lars Bergstrom

Researcher, Mozilla
Lars Bergstrom is a Researcher at Mozilla. He works on the Servo web browser engine and obtained his PhD in Computer Science from the University of Chicago. He has presented at research conferences (ICFP, IFL, USENIX-Windows) and has lectured both introductory and advanced systems... Read More →
MB

Mike Blumenkrantz

Mike Blumenkrantz is a Senior Engineer at Samsung R&D America. He is the maintainer of the Enlightenment desktop as well as a contributor to the Servo browser engine.


Friday June 5, 2015 14:00 - 14:50 JST
Mars

15:00 JST

GlusterFS - Overview & Future Directions - Haruka Iwao, 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

Friday June 5, 2015 15:00 - 15:50 JST
Orchid 3

15:00 JST

Scalable Private Cloud Storage with Full OSS Stack - Hiroshi Miura, NTT DATA
Cloud storage is now very popular for ordinal end users; there are many famous services such as Dropbox, Google drive, OneDrive and so on. Many organizations, that want to help employers or members keeping high productivity, are interested in cloud productivity toolset such as Google Apps, Zoho, etc. I'd love to introduce an original OSS combination; ownCloud, OpenStack Swift, Sheepdog and more for its purpose. This is introductory session for these OSSes.

Speakers
avatar for Hiroshi Miura

Hiroshi Miura

Manager, NTT DATA Corporation
Mr. Hiroshi Miura is an experienced speaker in areas of Linux, Python, OSS education and OpenStreetMap. He made presentations and panel sessions including LinuxCon 2010, PyConJP 2011, LinuxCon 2012, Enterprise User Meeting 2013. He is now providing proffessional services to customers... Read More →


Friday June 5, 2015 15:00 - 15:50 JST
Jupiter

15:00 JST

Meta-Debian: Extending Yocto Project's Poky - Kazuhiro Hayashi, Toshiba
For industrial products, kernel and open source software included in the system require fully customizable and longer term support. Distribution and its platform tools which satisfy such requirements are needed for developers to introduce Linux into embedded machines widely. Poky build system is well-designed to create Linux for embedded systems; users can fully customize Linux based on a default reference distribution provided by Poky. This presentation proposes an additional layer for Poky - meta-debian. meta-debian provides functions and meta data to generate customized Debian only by putting the layer into Poky build system without modifying original meta data. This is an example of Poky extension. Our purpose is to provide tiny and easy customizable Debian to embedded system because we'd like to use Debian's long term supported sources and make a contribution to that project.

Speakers
KH

Kazuhiro Hayashi

Software Engineer, TOSHIBA Corporation
Kazuhiro Hayashi works at TOSHIBA Corporation as a Software Engineer since 2010. The main part of his work is to develop Linux for various industrial embedded products. His another focus is to provide a common Linux distribution and its build infrastructure for effective product development... Read More →


Friday June 5, 2015 15:00 - 15:50 JST
Neptune

15:00 JST

OpenPOWER Technology Innovation - Paul Mackerras, IBM
The OpenPOWER Foundation, announced by Google, IBM, Mellanox, NVIDIA and Tyan in August 2013, is an open and global development alliance based on IBM's POWER microprocessor architecture.  Now, more than 110 members have joined and many innovative technologies are being developed. Linux and other open-source software plays a major role in the OpenPOWER ecosystem.  IBM has open-sourced the initialization and boot firmware for POWER8 machines, making it possible to construct a server-class machine where the whole software stack from the operating system down to the first
instruction executed at power-on, is entirely open source. This session will highlight two OpenPOWER technologies of particular interest to developers. The first is the integration of NVIDIA GPUs into POWER systems, together with CUDA integration provided by IBM Java.  The second, CAPI, allows us to integrate a Field Programmable Gate Array (FPGA)
into a POWER8 system, allowing computational functions programmed into the FPGA to be accessed directly from a user-space process without operating system overhead.  These technologies enable application developers to obtain outstanding performance at low cost for programs that handle large data arrays. These technologies are  already close to you for next generation Linux software and systems.

Speakers
PM

Paul Mackerras

Paul Mackerras works in the Linux Technology Center in IBM, where he leads a group of people whose mission it is to bring open-source virtualization to IBM's Power systems. He has been working on Linux since 1996, when he ported Linux to run on Apple Power Macintosh computers. He... Read More →


Friday June 5, 2015 15:00 - 15:50 JST
Soleil

15:00 JST

Stacked Vlan in Linux: with Report from Netdev 0.1 - Toshiaki Makita, NTT
Since 802.1ad vlan protocol was introduced in 2013, stacked vlan (double vlan tags) has been getting common on Linux. It can be used not only inside a data center network, but also in integrating Linux into Ethernet VPN, e.g. Metro Ethernet, which often consists of 802.1ad switches. However, there still remains a couple of challenges related to performance and usability. This presentation will give an overview of stacked vlan and its issues, and how they are being addressed. As this topic is discussed in the new Linux networking conference, Netdev 0.1, this presentation also gives a report of related topics on state-of-the-art technologies presented in Netdev.

Speakers
TM

Toshiaki Makita

Linux kernel engineer, NTT Open Source Software Center
Toshiaki Makita works for NTT Open Source Software Center (NTT OSSC), where he has been providing technical support for Linux kernel. He used to be a research and development engineer focusing on Ethernet VPN at NTT West, which is a regional carrier in NTT group. He has been an active... Read More →


Friday June 5, 2015 15:00 - 15:50 JST
Orchid 1 & 2