This event has ended. Visit the official site or create your own event on Sched.
October 23-26, 2017 - Prague, Czech Republic
Click Here For Information & Registration

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

CloudOpen Tracks [clear filter]
Monday, October 23

11:15 CEST

Automated Out-Of-Band Management with Ansible and Redfish - Jose Delarosa, Dell EMC
Ansible is an open source automation engine that automates complex IT tasks. It is a one-to-many agentless mechanism where complex deployment tasks can be controlled and monitored from a central control machine.

Redfish is an open industry-standard specification and schema designed for modern and secure management of platform hardware. On Dell EMC PowerEdge servers the Redfish management APIs are available via the iDRAC, which can be used by IT administrators to easily monitor and manage at scale their entire infrastructure using a wide array of clients.

Together, Ansible and Redfish can be used by system administrators to fully automate at large scale server monitoring, provisioning and update tasks from one central location, significantly reducing complexity and helping improve the productivity and efficiency of IT administrators.

avatar for Jose Delarosa

Jose Delarosa

Senior Linux Engineer, Dell EMC
Jose is a Senior Linux engineer at Dell EMC. He spends most days keeping things from breaking, helping others, learning new things and keeping customers happy. Current projects include Ceph, Docker, Ansible and various Linux system administration responsibilities. Past speaking experiences... Read More →

Monday October 23, 2017 11:15 - 11:55 CEST

11:15 CEST

How to Automate a Large Scale Government OpenStack Private Cloud - Kenneth Tan & Mihaela Constantinescu, Sardina Systems
This talk will draw on lessons learnt through the full lifecycle of large scale government OpenStack private cloud, from design and implementation, through to operation and upgrade. How to bring into operation a cost effective, efficient, OpenStack environment, which scales to span multiple geographical locations? Is it possible to keep the OpenStack cloud service up and running when downtime is not an option? At the same time, how can we track OpenStack versions, without impacting the service at all?

avatar for Dr. Kenneth Tan

Dr. Kenneth Tan

Director, Sardina Systems
Dr Kenneth Tan has been in the large scale systems sector for 18+ years. At Sardina Systems, Kenneth leads the product development team responsible for building OpenStack automation software, to optimize OpenStack cloud operations. He was previously with CloudFabriQ, BNP Paribas... Read More →

Monday October 23, 2017 11:15 - 11:55 CEST

12:05 CEST

Multi-Repo, Multi-Node Gating at Massive Scale - Monty Taylor, Red Hat
The OpenStack Infra team runs one of the world's largest Open Source CI/CD systems in service of OpenStack's early decision to mandate that all merges only be performed by automation if and only if all tests pass.

Zuul is the engine developed to handle this, but it is not OpenStack specific. With the rise of microservices and kubernetes, the number of multi-repo projects is increasing, as is the need for CI systems that understand them. The most recent version of Zuul has been reworked to make it easy for other people, communities or organizations to harness its power regardless of any relationship with OpenStack.

We'll talk about the things that make Zuul special - multi-repository dependencies, optimistic branch prediction and deep Ansible integration. And we'll walk through how to get started with a private or a public Zuul.

avatar for Monty Taylor

Monty Taylor

Red Hat

Monday October 23, 2017 12:05 - 12:45 CEST

12:05 CEST

Scaling Massive, Real-Time Data Pipelines with Go - Jean de Klerk, Pivotal
Big data often means huge processing pipelines. Passing 10,000 messages a second between apps and datastores is tough. Add on filtering and parsing and you’re bound to run into problems. At Pivotal Cloud Foundry, we have to handle several factors of that load each day, which means our programs need to be highly concurrent and very careful with state. Let’s look at some patterns and best practices that can be taken advantage of in order to handle this load.

avatar for Jean de Klerk

Jean de Klerk

Senior Software Engineer, Pivotal Labs
Jean is a senior software engineer working on Pivotal's open source Cloudfoundry PaaS, building highly concurrent microservices for metrics, logs, and autoscaling. Jean's open source experience includes work on projects such as Pivotal UI, the React.js and Spring projects, several... Read More →

talk key

Monday October 23, 2017 12:05 - 12:45 CEST

14:20 CEST

Databases in the Hosted Cloud - Colin Charles, Percona
Today you can use hosted MySQL/MariaDB/Percona Server/PostgreSQL in several "cloud providers" in what is considered using it as a service, a database as a service (DBaaS). Learn the differences, the access methods, and the level of control you have for the various public cloud offerings:
- Amazon RDS including Aurora
- Google Cloud SQL
- Rackspace OpenStack DBaaS
- Oracle Cloud's MySQL Service

The administration tools and ideologies behind it are completely different, and you are in a "locked-down" environment. Some considerations include:
* Different backup strategies
* Planning for multiple data centres for availability
* Where do you host your application?
* How do you get the most performance out of the solution?
* What does this all cost?
* Monitoring

Growth topics include:
* How do you move from one DBaaS to another?
* How do you move all this from DBaaS to your own hosted platform?

avatar for Colin Charles

Colin Charles

Consultant, GrokOpen
Colin Charles is the Managing Consultant at GrokOpen. Previously, Colin was on the founding team of MariaDB Server, worked at MySQL and Percona, and worked actively on the Fedora and OpenOffice.org projects. Colin has been a MySQL user since 2000. He’s well known within open source communities, enjoys building business and market entry in APAC and has spoken at many conferences... Read More →

Monday October 23, 2017 14:20 - 15:00 CEST

14:20 CEST

Introduction to The Foreman - Lukáš Zapletal, Red Hat
The Foreman is a complete lifecycle management tool for physical and virtual servers. It’s a great tool for handling your infrastructure - from simple networks to multi-thousand node setups across multiple data centers. This talk will cover the basics of The Foreman and some of it’s popular plugins.

We’ll see Foreman in action:
* automatically identifying an unknown server on the provisioning network with the discovery plugin
* provisioning the host and configuring puppet parameters for it
* executing commands on a list of hosts with the remote execution plugin

avatar for Lukáš Zapletal

Lukáš Zapletal

Software Engineer, Red Hat
The Foreman open-source project core member with focus on hardware discovery, bare-metal provisioning, non-Intel architectures, PXE and SELinux. Works in Red Hat Satellite 6 engineering team.

Monday October 23, 2017 14:20 - 15:00 CEST

15:10 CEST

Monitoring At Scale: What was Recently Done and What’s Next in oVirt - Arik Hadas, Red Hat
Monitoring the state of high number of running virtual machine instances is crucial for effective management of large-scale virtual environments. It provides invaluable information for system administrators about the health of the environment as well as the required information for automatic processes that are executed by a management system, such as automatic restart and load balancing of virtual machines. Doing that efficiently is important because it is a periodic task running in the background that may grab a lot of resources, thus potentially affect the performance and user experience of the management system. Consequently, improving virtual machines monitoring was at the focus of recent versions of oVirt. In this talk, Arik Hadas will present what has been done recently in oVirt in order to improve virtual machines monitoring and provide a glimpse into the on-going work in that area.

avatar for Arik Hadas

Arik Hadas

Principal software engineer, Red Hat
Arik is a principal software engineer with more than a decade of experience in contributing to open source projects. He is a co-maintainer in the oVirt project and loves sharing his technical experience. Check out his blog at http://ahadas.github.io... Read More →

Monday October 23, 2017 15:10 - 15:50 CEST

15:10 CEST

Overcoming Logging Challenges for Cloud Native Applications at Scale - Eduardo Silva, Treasure Data
There are tried and true methods for debugging and monitoring standalone applications; however, applications at scale are a different story. There are huge complexities involved that affects how log management needs to be handled to perform a safe data analysis. Logging becomes challenging in environments with distributed applications and/or microservices as new patterns and specialized tools are required.

The following presentation will dive into the complexity of log management at scale, the internal phases of a logging pipeline, constraints in containers, and approaches like Log Forwarders and Log Aggregators that can be used to achieve a scalable logging layer in a Kubernetes cluster.

avatar for Eduardo Silva

Eduardo Silva

Principal Engineer, Arm Treasure Data
Eduardo is a Principal Engineer at Arm Treasure Data, he is the author and maintainer of Fluent Bit Log Processor, a CNCF sub-project under the umbrella of Fluentd. He is an international speaker in Open Source conferences, he has participated in Scale California, LinuxConf AU, Linux... Read More →

Monday October 23, 2017 15:10 - 15:50 CEST

16:20 CEST

Architecturing and Securing IoT Cloud Platforms - Drasko Draskovic & Dejan Mijic, Mainflux
IoT device management and multi-protocol messaging platform demands specific architectural decisions and high-concurrency approaches due to the massive number of devices we expect to be connected in the near future. All of this must be done with high security in mind, as well as respecting requirements for low-power, memory constrained devices with intermittent connectivity and limited bandwidth.

We will analyze the architecture, implementation and testing procedures needed for creating an industry-grade IoT platform and propose a patent-free open-source solution based on a set of scalable containerized microservices, with high concurrency and high security based on PKI with specific ciphers and encryption procedures that are suitable for constrained devices.

Similar tutorial session was accepted and presented on Open Networking Summit 2017 in Santa Clara and was very well received.

avatar for Drasko DRASKOVIC


CEO, Mainflux
Drasko is an IoT expert with over 15 years of professional experience. He hacked embedded Linux SW and HW device drivers, designing complex wireless systems in telecom industry: he was working on OMAP platform in Texas Instruments, designed 4G multi-protocol femto-cells in Alcatel-Lucent... Read More →
avatar for Dejan Mijic

Dejan Mijic

Software Architect, Mainflux
Software engineer interested in distributed systems. Co-architect and principal developer of Mainflux IoT platform. PhD candidate at the University of Novi Sad, Serbia.

Monday October 23, 2017 16:20 - 17:00 CEST
  CloudOpen Tracks

16:20 CEST

Beyond Configuration Management -- Event-Driven Linux Systems Management with SaltStack - Mike Place, SaltStack
With the rise of containerization and microservice architectures, what it means to manage Linux systems in production is shifting rapidly. Beyond simply deploying systems into production, how do we observe those systems, react to changes and manage the complete lifecycle of Linux instances from deployment to de-provisioining? In this presentation, the maintainer of the SaltStack project -- Mike Place -- will show us how adopting an event-driven approach to the management of Linux systems can break down the walls between monitoring and configuration management in order to design automation workflows that keep systems operating optimally day and night.

avatar for Mike Place

Mike Place

Open Source Manager, SaltStack
Mike Place is the manager of the open-source team at SaltStack and the maintainer of the Salt project for systems management -- one of the most actively developed Python projects in the world.

Monday October 23, 2017 16:20 - 17:00 CEST

17:10 CEST

EulerOS Isula: Born For Cloud, Drive Cloud Native - Wei Xiong, Huawei

Almost all of huawei’s business are moving to cloud native fastly, The OS team is facing very different scenarios than before, wide variety of devices, container based applications, serverless service, IOT infrastructure, edge computing and so on, different scenarios may bring up different OS requirements, How does OS team to handle those differences? What’s OS’s development trend? What technology may bring up significant changes to OS area? Huawei’s OS team try to answer those questions by isula project. I this session, we present the general idea of Isula project and its rough architecture.

avatar for Dr Wei Xiong

Dr Wei Xiong

OS architect, Huawei
Xiong Wei, joined Huawei in 2014, is now the 2012 laboratory Central Software Institute server operating system chief architect, openEuler technical committee member; Nankai University, doctor of engineering, in TurboLinux, WindRiver and other companies as R & D person in charge... Read More →

Monday October 23, 2017 17:10 - 17:50 CEST

17:10 CEST

OpenSDS and Storage Transformation for Cloudification at Vodafone - Cosimo Rossetti, Vodafone & Steven Tan, Huawei
OpenSDS is an open-source project created to address the storage challenges, particularly in scale-out cloud native environments, with heterogeneous storage platforms. The project promotes the use of simplified storage interfaces using a scalable storage controller architecture with open standard APIs, with the objective of providing application-oriented storage services. 

As part of its IT infrastructure modernization for cloud, Vodafone is going through a storage transformation adopting Software-Defined Storage (SDS) strategy for their Software-Defined Data Center (SDDC) evolution.  In this presentation, Steven Tan, Chief Architect for Storage Management and Storage-as-a-Service at Huawei, will be introducing OpenSDS and the storage challenges the project is addressing; and Cosimo Rossetti, Lead Storage Architect at Vodafone will be sharing their software-defined storage strategy, and how they hope to leverage OpenSDS in their storage transformation.


Cosimo Rosetti

Architect, Vodafone
Cosimo Rossetti is helping to define Global Target Architecture and Strategy and build a modern storage infrastructure at Vodafone as Storage Lead Architect. Prior to Vodafone, he worked in several service providers companies for 15+ years as Pre-Sale Technical Solution Architect... Read More →
avatar for Steven Tan

Steven Tan

SODA Foundation Chair, VP & CTO Cloud Solution - Storage, SODA/Futurewei
Steven Tan is VP & CTO Cloud Solution, Storage at Futurewei where he is responsible for open source data management and storage strategy and partnerships. He brought global leaders together, and founded the SODA Foundation (based on the OpenSDS project) which he currently serves as... Read More →

Monday October 23, 2017 17:10 - 17:50 CEST
Tuesday, October 24

10:55 CEST

Everything You Always Wanted to Know About Object Storage - Orit Wasserman, Red Hat
The rapid growth of unstructured data is fueling the need for a next generation storage that’s flexible, economical, and scalable enough to handle the petabytes of data being created every day. Object storage is the answer!

Ceph is a highly available distributed software defined storage, that provides two object storage interfaces:
  • Rados provides native object storage API using a rich library with C/C++, java, python, go and several others bindings.
  • Ceph RGW (Rados Gateway) provides HTTP REST API that is Amazon S3 and openstack swift compatible.
In this talk I will introduce object storage foundations, best practices and Ceph object storage solution.

avatar for Orit Wasserman

Orit Wasserman

Senior Principal Software Engineer, Red Hat
Orit is a senior principal software engineer at Red Hat, focusing on Container and multi cloud storage. She was a principal architect at Lightbits labs working on NVMe/TCP software-defined storage. At Red Hat, she worked on Ceph object storage (Ceph Rados Gateway), a highly available... Read More →

Tuesday October 24, 2017 10:55 - 11:35 CEST

10:55 CEST

PNDA: Learn to Bootstrap an Open Source Big Data Platform - Jeremie Garnier, Cisco DevNet
Innovation in the big data space is rapid, but combining multiple technologies into an end‐to‐end solution is extremely complex and time consuming. The vision of PNDA, the open source big data analytics platform, is to remove this complexity and allow developers to focus on their applications and business outcomes. PNDA brings together a number of open source technologies to provide a simple, scalable big data analytics platform to achieve this. Attendees to this session will gain a better understanding of a Big Data reference architecture and learn to bootstrap a platform within their own environment.


Jeremie Garnier

Development Expert, Cisco DevNet
Jérémie has been working in Cisco’s Innovation Department for over ten years, focusing on Cloud software solutions including Solr/Lucene, Kafka, Zookeeper, Hadoop, as well as Android and HTML5. While this role has afforded the opportunity to learn and stay on the cutting-edge... Read More →

Tuesday October 24, 2017 10:55 - 11:35 CEST

11:45 CEST

Cloud + Edge, an Approach Using Kubernetes - Yin Ding, Huawei
With IoT and mobile devices, more and more data are being created at the edge of the network, quite a lot of them should be processed at the edge instead of sending over the network to a central cloud to process. Cloud+Edge will be a more efficient architecture for this modern world.

In this session we will discuss our approach using Kubernetes to build our Edge cloud and interact with our central cloud to implement a more efficient Cloud+Edge architecture. The topics include deployment, resource orchestration, message passing, function distribution etc.


Yin Ding

Chief Virtualization Technologist, FUTUREWEI TECHNOLOGIES, INC.
Dr. Yin Ding is the new generation Virtualization Technologist at Huawei Cloud BU. Yin is responsible for creating and communicating technical vision and strategy for Cloud Infrastructure and Container Technology business. He works closely with R&D teams to bring the new generation... Read More →

Tuesday October 24, 2017 11:45 - 12:25 CEST

11:45 CEST

Pipeline as Code For Your Infrastructure as Code - Kris Buytaert, Inuits.eu
Infrastructure as Code has been a native open source innovation happening over the past decade, with languages such as Puppet and Chef growing fast ops folks have been adopting development practices to run their infrastructures. In the mean time the operations people together with the developers ran into the limitations of traditional CI platforms where a lot of "Dirty Clicking" was the norm , creating a pipeline for a stack, or new project was a boring manual experience. This talk will document the use of Pipeline as Code (Jenkins DSL) to generate a Containerized deployment and test infrastructure testing different versions of Puppet code (actually this also serves as a Puppet 3 to Puppet 5 migration) for different customers. (Albeit we did exaclty the same for Python, PHP and Ruby stacks..)

avatar for Kris Buytaert

Kris Buytaert

Chief Yak Shaver, Inuits.eu
Kris Buytaert is a long time Linux and Open Source Consultant. He's one of instigators of the devops movement, currently working for Inuits He is frequently speaking at, or organizing different international conferences He spends most of his time working on bridging the gap between... Read More →

Tuesday October 24, 2017 11:45 - 12:25 CEST
  CloudOpen Tracks

14:05 CEST

Enterprise Grade Security and Compliance Using Apache Atlas and Apache Ranger - Vimal Sharma, Hortonworks
With an ever increasing need to secure and limit access to sensitive data, enterprises today need an open source solution. Apache Atlas - which is the metadata and governance framework for Hadoop joins hands with Apache Ranger - security enforcement framework for Hadoop to address the need for compliance and security. Vimal will discuss the security and compliance requirements and demonstrate how the combination of Atlas and Ranger solves the problem. Vimal will focus on Tag based policy enforcement which is an elegant solution for large Hadoop clusters with wide variety of data

avatar for Vimal Sharma

Vimal Sharma

Software Engineer, Hortonworks
Vimal Sharma is Apache Atlas PMC and Committer at Hortonworks. Vimal is highly passionate about Hadoop stack and has previously worked on scaling backend systems at WalmartLabs using Spark and Kafka. Vimal was a speaker at ApacheCon BigData 2017 where he spoke on Metadata governance... Read More →

Tuesday October 24, 2017 14:05 - 14:45 CEST
  CloudOpen Tracks
  • Experience Level Any

14:55 CEST

Scaling YARN To 10K+ Nodes -- The Challenge - Atri Sharma, Microsoft
The talk shall focus on the scalability challenges around YARN and the way they have been solved at Azure Data Lake. The talk shall go into details of the federated model that has been introduced and how that allows massive scalability with top notch performance.

avatar for Atri Sharma

Atri Sharma

SDE-II, Microsoft
A distributed systems engineer, committer on Apache Apex, PMC Member on Apache MADLib, PPMC Member on Apache HAWQ and major contributor in PostgreSQL Project, having implemented GROUPING SETS, ROLLUP, CUBE and Ordered Set Aggregates

Tuesday October 24, 2017 14:55 - 15:35 CEST

14:55 CEST

Containerized Full-Stack Rust for IoT - Claus Matzinger, Microsoft
Rust's ability to work on many different platforms across the full stack is great, but how easy is it to use our favorite language from data collection to storage? Can a compiled Rust application work well in containers across CPU architectures and libc implementations? This talk will show challenges and solutions to implementing a containerized client-server application to collect images and IoT data, including but not limited to: Testing, serialization, CI, cross-compilation, drivers, and coroutines. 

avatar for Claus Matzinger

Claus Matzinger

Technical Evangelist, Microsoft
Now at Microsoft, Claus used to run field engineering at Crate.io, the creators of CrateDB, a distributed SQL database. As a former CTO of a health startup and an Raspberry Pi and Rust (the programming language) enthusiast, Claus also maintains several drivers for sensors in the Rust... Read More →

Tuesday October 24, 2017 14:55 - 15:35 CEST

16:05 CEST

Multi-Cloud Infrastructure Management by Infrakit - Yuji Oshima, NTT Labs
Due to the generalization of the cloud, some people today try to use multi-region and multi-cloud environment more and more.
These systems have three merits:
1. to improve availability
2. to avoid cloud lock in
3. to optimize the cost
That is, even when you have some problems with one cloud, you can use another cloud as a substitution.
In addition, you can change from one to another easily, so you don’t need to be afraid of lock-in.
Finally, using multi-region and multi-cloud helps you optimize the whole cost.
For example, AWS and GCP have various instance sizes and types such as Spot instance and Preemptible VM, and prices are different for each of sizes or types.
According to the change in the circumstances, you can choose and allocate such instances properly using multi-region and multi-cloud systems.

In spite of these potential advantages, you don’t know how to use effectively multi-region and multi-cloud yet.
There are two main reasons in this:
1. It is too complicated to define infrastructure across several cloud providers.
2. It is too difficult to scale infrastructure if you want to choose reasonable providers and properties.
Terraform can solve one of the problem.
Terraform is a useful tool for handling the former problem and help your deployment.
However, management of infrastructure after deployment is beyond their scope.
I would like to propose a new multi-cloud operation using Infrakit, Docker’s software.
Infrakit will solve both problems all at once.
In other words, you can freely add custom rules for scaling and modify infrastructure without being disturbed by stopping service.
First, I will talk about how to define and deploy infrastructure with Infrakit.
Then I will describe the implementation of infrastructure management rules and the way to add custom rules.
Finally, I will demonstrate that you can deploy and scale infrastructure across cloud providers and on-premice environment with Infrakit.

avatar for Yuji Oshima

Yuji Oshima

Researcher, NTT
Yuji Oshima is a researcher working for NTT Labs. His group has been developing open source software such as Ryu(SDN controller), GoBGP (software BGP router), and Sheepdog (distributed storage system for QEMU), and contributing to kubeflow, etcd, and docker engine. He is one of owners... Read More →

Tuesday October 24, 2017 16:05 - 16:45 CEST
  CloudOpen Tracks

16:05 CEST

Tutorial: Hands-on with Containerized Deployment of OpenStack - Charles Eckel, Cisco
This lab is part lecture and part hands-on, with exercises being used to drive home concepts covered in the lecture.

Hearing a lot about OpenStack and want to check it out for yourself? See how quick and easy it is to install start using OpenStack using containers running within a VM on your own laptop. OpenStack Kolla provides production ready tools to deploy OpenStack services as Docker containers that can be managed and upgraded easily. To help you explore Kolla and OpenStack, we provide access to a VM with a containerized deployment of OpenStack and step by step instructions. Acquaint yourself with the environment. Learn your way around Horizon (GUI) and the CLI to view and operate your OpenStack cloud. Best of all, take what you learn with you and experiment on your own to discover all OpenStack offers you.

For the best experience, bring a laptop with at least 16GB of RAM. We will be using VirtualBox. Install it plus the extension packs prior to the tutorial to save some time. The tutorial makes use of a Learning Lab hosted in Cisco DevNet. To get a jump on things, you can:

1) Sign up for DevNet free and easy using this event specific URL https://developer.cisco.com/join/oss-eu17

2) Access the Learning Lab at https://learninglabs.cisco.com/lab/openstack-install/step/1
3) Follow the instructions in the “How To Setup Your Own Computer” section, including downloading the 7 GB OVA file
I will bring copies of this file on some USB thumb drives, but best to download ahead of time if possible.

Follow me on Twitter @eckelcu to see any last minute updates. See you soon!

avatar for Charles Eckel

Charles Eckel

Principal Engineer, Global Technology Standards, Cisco
Charles is a recognized champion of open source, standards, and interoperability. As a member of Cisco's Global Technology Standards team, Charles is responsible for identifying and guiding open source efforts related to key standards initiatives. In IETF, he started and runs the... Read More →

Tuesday October 24, 2017 16:05 - 17:35 CEST

16:55 CEST

Open Source Cloud Ecosystem for Next-Gen Disaggregated Datacenters - Andrea Reale, IBM Research
Modern computer systems have been for long designed as monolithic blocks built around the boundaries of the motherboard. This talk looks at novel architectures for low-power and high-utilization disaggregated Cloud datacenters that break those boundaries. Andrea Reale will give an overview of the hardware and software enabling the dynamic creation of fit-for-purpose computing environments from a pool of disaggregated resources. He will focus on the open source blocks needed to build a disaggregate Cloud, including the Linux kernel, Qemu+KVM, Openstack and JanusDB. He will present how IBM Research is building a proof-of-concept disaggregated Cloud by exploiting and extending these opensource components. A tiny-scale live demo of the prototype will be brought on stage, demonstrating how remote disaggregated memory can be allocated and used transparently from application binaries.

avatar for Andrea Reale

Andrea Reale

Research Engineer, IBM Research
Andrea Reale is a Research Engineer in the High Performance Systems team at IBM Research in Ireland. His interests and work focus around middleware and OS-level support for high performance distributed systems. Until early 2017, he has been a contributor and founding PMC member for... Read More →

Tuesday October 24, 2017 16:55 - 17:35 CEST
Wednesday, October 25

11:15 CEST

GPGPU on OpenStack - The Best Practice for GPGPU Internal Cloud - Masafumi Ohta, Japanese Raspberry Pi Users Group
GPGPU on OpenStack - the best practice for GPGPU internal cloud ( Masafumi Ohta, Itochu Techno Solutions) - GPGPU on OpenStack is one of the OpenStack use cases automotive companies may use it as huge temporary instances and trials for their developments - Machine Learning, HPC and more like Amazon EC2 as internal cloud but it hasn’t been documented yet in detail especially for 'automotive IT users'. In this session, Masafumi will review the backgrounds why GPGPU is needed for virtualization, the detailed mechanism, and settings how to use GPGPU on virtualization environments, discuss the challenges we face with on GPGPU on OpenStack open-source based virtualization cloud system, the issues on GPGPU on virtualization systems, the roadmap and feature may address those challenges and issues and the good use case automotive companies might adopt for their business.

avatar for Masafumi Ohta

Masafumi Ohta

Founder and Representative, Japanese Raspberry Pi Users Group
Masafumi is leading Raspberry Pi community in Japan and volunteering Raspberry Pi Foundation from farthest east country, Japan,.He has helping their business and encourage Raspberry Pi related projects with Raspberry Pi Foundation.Masafumi has elected ARM INNOVATOR by ARM+Hackster.io... Read More →

Wednesday October 25, 2017 11:15 - 11:55 CEST
  CloudOpen Tracks
  • Experience Level Any

12:05 CEST

Dirty Clouds Done Dirt Cheap - Matthew Treinish, IBM
OpenStack is an open source cloud ecosystem that is designed to work well at all scales. But, how does it work for an individual wanting to deploy a cloud with their own resources?

I've wanted to build a small cloud at home for some time, mostly to serve as a testbed for my development on OpenStack, but also as a platform to virtualize my growing home infrastructure. So I allocated a budget of ~$1500 USD (the price of my first desktop computer) and set out to build a cloud.

This talk will provide an overview of my experience building a small compute cloud from scratch, on essentially a shoestring budget; from acquisition of hardware, through installation and configuration of the cloud, to my use cases for the cloud. It will provide an overview of building a minimum compute cloud by hand, and how to get started using a personal cloud and potential future applications.

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 →

Wednesday October 25, 2017 12:05 - 12:45 CEST
  CloudOpen Tracks

14:15 CEST

Don't Break Production - Nigel Babu, Red Hat
Gluster is a project that provides software-defined storage for a wide range of use-cases. Given the wide range of options, there are some combinations of options which we break during an upgrade inadvertently. We also tend to do a lot of work for use cases that aren't very common as well. In the last year, we've been working towards refining our focus to make better use of our time developing the project and adapting specific use cases that we will officially support. This talk will lay down what we learned in the process and how your project can adopt these values as well.

avatar for Nigel Babu

Nigel Babu

Sysadmin, Red Hat
Nigel is a developer turned sysadmin, who now does CI Infrastructure for the Gluster project at Red Hat. In the past, he's spoken at Linuxcon Berlin, Pycon Pune, and Open Source Summit Tokyo.

Wednesday October 25, 2017 14:15 - 14:55 CEST

15:05 CEST

Self-Healing Networkings: Responding to Your Network in Real Time - Tyler Christiansen, Sauce Labs
Self-healing infrastructure and automated remediation are hot topics in today's world, but they're rarely discussed in the context of networks. When they are, it's rarely in an open source--or even practical--manner. In this presentation, Tyler Christiansen will apply the concepts of automatic validation of changes and real-time event response to traditional network devices. This will include an exploration of what is both possible and practical for organizations to adopt using only open source tools that are highly extensible. The presentation will also discuss some of the current challenges faced by network operators today and how event-based reactions can help. Live demonstrations will be used to illustrate key areas.


Tyler Christiansen

Network Architect, Sauce Labs
Tyler Christiansen is a network engineer with experience across a plethora of verticals, including carrier core IP/MPLS, international ad serving, and network software development. He has worked in public, private, and hybrid clouds in a variety of roles, including Systems Engineer... Read More →

Wednesday October 25, 2017 15:05 - 15:45 CEST
  CloudOpen Tracks
  • Experience Level Any

16:15 CEST

Integration of Flexible Storage with the API of Gluster - Niels de Vos, Red Hat
No matter where an application is running, it will most likely need some form of storage. Gluster is a distributed scale-out filesystem that offers interfaces for many different use-cases and workloads.

During this talk Niels will show how different applications addressed their storage needs by integrating with the API that Gluster provides. Examples of mature integrations are included in Samba, QEMU and NFS-Ganesha. Bindings for different languages (C, Golang, Python, Java and others) are available. This enables many other projects to adapt their storage backends for Gluster. In this session the attendees will learn how to get started te integrate Gluster in their own applications.

avatar for Niels de Vos

Niels de Vos

Senior Software Engineer, Red Hat
Niels is a core-developer and maintainer for Gluster. He is employed by Red Hat and works together with other teams who provide professional support for Red Hat Gluster Storage. The main areas where Niels is active include network protocols, low-level/Operating Systems improvements... Read More →

Wednesday October 25, 2017 16:15 - 16:55 CEST

17:05 CEST

Building Robust Streaming Data Pipelines with Apache Spark - Zak Hassan, Red Hat
There are challenges to architecting a solution that will allow for developers to stream data into Kafka and be able to manage dirty data which is always an issue in ETL pipelines. I'd like to share lessons learned and demonstrate how we can put Apache Kafka, Apache Spark and Apache Camel together to provide developers with a continuous data pipeline for the Spark applications. Without data it is very difficult to take advantage of its full capabilities of Spark. Companies sometimes have their data stored in many different systems and Apache Camel allows developers to Extract, Transform and Load their data to many systems Apache Kafka is one example. Apache Kafka is great for aggregating data in a centralized location and Apache Spark already comes with a built in connector to connect to Kafka. I'll also be explaining lessons learned from running these technologies inside docker.

avatar for Zak Hassan

Zak Hassan

Senior Software Engineer - AI/ML CoE, CTO Office, Red Hat Inc.
Currently focused on developing analytics platform on OpenShift and leveraging Open Source ML Frameworks: Apache Spark, Tensorflow and more. Designing high performance and scalable ML platform that exposes metrics through cloud-native technology: Prometheus and Kubernetes.

Wednesday October 25, 2017 17:05 - 17:45 CEST