Sorry, you need to enable JavaScript to visit this website.
Home / OpenStack* on Intel® Architecture / OpenStack* on Intel® Architecture - OpenStack Summit Austin

OpenStack Summit Austin

Published by:
Last modification:
Apr 07, 2021

Join Intel at OpenStack Summit Austin FOR A FULL ROUND-UP!

We're looking forward to catching up with you in Austin. Please join us for a variety of engaging sessions:

-- Keynote by Imad Sousou, Vice President & General Manager of Intel's Open Source Technology Center 

-- Over 40 sessions & 8 demos across compute, networking, storage and security (Intel Booth B23)

-- Learn more about the OpenStack Innovation Center (OSIC) in the Design Summit Developer Lounge

-- Chance to win a cool mini-rack in the "Bounty for Big Ideas" developer contest


Command Presence Workshop Recognized by the Anita Borg Institute and Harvard Business Review, the Command Presence Workshop is intended to teach participants how to handle senior-level, high-pressure meetings. The majority of the workshop is a task force convened to address a simulated, but very realistic crisis event.  In the class, each participant assumes a role and is given about 30 minutes to prepare 1-2 slides. Once the class has prepared, a role-play simulation begins. The workshop participants will experience rapid-fire questioning, interruption, and other situations that occur in senior-level meetings, but in a safe environment conducive to learning techniques for handling such situations. It is not uncommon to hear occasional gasps of disbelief from the participants! Ruchi Bhargava (Intel), Malini Bhandaru (Intel), and a panel of leaders representative of the OpenStack leadership in the community Sunday, April 24, 09:30-12:30
Architect Application Workloads for OpenStack Clouds The Enterprise Working Group has been discussing various strategies to help enterprise architects understand how to deploy tenant-level workloads on OpenStack. With the Project Navigator and Sample Configuration in place, the team has started creating multiple Workload Reference Architecture diagrams targeted for different use cases. In this session, we will present our work that has been done over the past few months and share reference architectures to the audience. Craig Sterrett (Intel), Megan Rossetti (Walmart) Monday, April 25, 11:15-11:55
Let's build an OpenStack Cross Project and User Story Dashboard! The OpenStack community currently has several ways to track requirements and implementation at the project level, from blueprints to specs to Gerrit. However, it is still fairly difficult to track cross-project needs and user stories (concepts) that require work across multiple projects and, potentially, multiple releases. In this session, we will cover the need for greater visibility from the Product WG that aligns well with the general cross-project development tracking needs in the community, which resulted in our project to create a cross-project user story dashboard. We have started working on a prototype, and will provide a demonstration of our progress thus far. Shamail Tahir (IBM), Marcela Bonell (Intel), Victor Diaz (Intel) Monday, April 25, 11:15-11:55
Meet the OpenStack Personas (so far) In projects with many users, personas clarify users, goals, and the effects of decisions about the backend, messages, APIs, CLIs, and GUIs on the many different users. The OpenStack UX project and OpenStack product group hosted a workshop at the IBM Design Austin Studio to develop personas that represented a wide variety of OpenStack consumer scenarios. Attendees included IBM, HPE, Intel and RackN. The team, along with the OpenStack Foundation, conducted a study in 2016 to validate the new personas, which included: Infrastructure Architect, Service Admin, Cloud Ops, Customer Ops, and DevOps Development. In this session, we will review the results of this work and discuss our vision with an example. Attendees will meet the personas and learn about the contributions personas can make to the community. We also welcome your input about the personas and our vision. Pieter Kruithof (Intel), Jeffrey Calcaterra (IBM), Stan Li (IBM) Monday, April 25, 11:15-11:55
Intelligent Workload HA in OpenStack Although Openstack natively supports some forms of high availability (HA), there is a big gap from the enterprise perspective. One of these gaps is live migration of VMs. In this session, we explore the development of an HA framework that is both reactive and proactive to address this gap. With reactive HA, we will show how the VMs running on a compute node that goes down can automatically be restored. With proactive HA, we will show how the framework triggers live migrations in cases of high thermal signature, predictive hardware failures, host maintenance mode, slow application performance due to bottlenecks, and other custom triggers (ceilometer alarms, zabbix) thus mitigating VM downtime for enterprise workloads. Nishi Ahuja (Intel), Pramod Bhandiwad (Tata Communications), Prateek Goel (Tata Communications), Jason Venner (Mirantis) Monday, April 25, 12:05-12:45
Application Ecosystem Work Group Discussion If you’re an enterprise architect, cloud operator, application developer or business/marketing member of the community who is interested in helping OpenStack to be the leading cloud solution in the Enterprise IT market segment, please come join this discussion on what the barriers are for deployment today and how we can remove them. Tell us what the roadblocks, issues, or requirements are of bringing OpenStack cloud into your environment. Victor Diaz (Intel), John Callaghan (HP) Monday, April 25, 14:00-14:40
Lessons Learned in Building & Operating the World's Largest OpenStack Dev Cloud Does your code scale? Come learn what it took to deploy the world's largest developer test cloud through the OpenStack Innovation Center (OSIC), and how we operate and maintain this environment. From architects who designed the clusters and operators who manage them, gain insights into best practices and lessons learned, and take away operational advice to deploy OpenStack at scale. TBC Monday, April 25, 14:15-14:45
Swagger as a Standard: A New API Definition Way for Doc & Imp Swagger, defined by the OpenStack API-WG, is now a standard, providing guidelines for API documentation for all OpenStack projects asa community. As a result, projects use common tooling, consistent outlines, and study exemplary examples. Swagger is not only a standard for API documentation but also for client side implementation. By getting Swagger data via REST API from the server side, clients can know available APIs and usage of these APIs, which aids OpenStack interoperability. We are now planning to implement Swagger in Nova and Magnum. On the documentation side, we are implementing a tool, "fairy-slipper," for migrating current API documentation to Swagger format. This session provides an overview of Swagger, how to apply Swagger to each project, and merit for clients. Alex Xu (Intel), Ken'ichi Ohmichi (NEC), Motohiro Otsuka (NEC), Anne Gentle (Rackspace) Monday, April 25, 15:40-16:20
Deep Dive Into Neutron Upgrade Story Everyone agrees that cloud upgrades must be smooth, quick, and reliable, with no downtime on data nor control plane. From the technical point of view, this is a big challenge, especially for Neutron which is well known as one of the most complex projects in OpenStack. In Mitaka, a dedicated Neutron Upgrades team was established to track and push forward upgrade items. In this session, we will show why it's important for developers and operators to understand all kinds of upgrade process implications to keep their users happy. We'll cover what was achieved by the Neutron Upgrades team in the Liberty and Mitaka cycles, starting from agent side work to the latest and greatest object versioning pieces that pave the way for more reliable rolling upgrades. We'll explain the challenges that we faced and the solutions we found so that other projects can learn from our trials and mistakes. Finally, we will list what's currently being worked on, and we'll glimpse the rolling upgrades roadmap. Artur Korzeniewski (Intel), Ihar Hrachyshka (Red Hat), Rossella Sblendido (SuSe) Monday, April 25, 16:40-17:20
Lockless Upgrades: How to Upgrade OpenStack Without Locking Up Your DB SQL databases require updating the schema to store new data structures; however, many DDL operations lock the tables. How can we change the schema while keeping our services online? What are the alternatives? Grzegorz Grasza (Intel), Alicja Kwaśniewska (Intel), Roman Dobosz (Intel) Monday, April 25, 17:30-18:10
OpenStack Development & Testing at Unprecedented Scale Through the OpenStack Innovation Center (OSIC), an environment is now available for use by the OpenStack community to develop and test features, functionality and code at scale never before possible. This environment, comprised of two 1,000-node clusters, represents the world’s largest developer test cloud for OpenStack. It offers a nexus for collaboration across the community, open to those committed to advancing the OpenStack platform. Join us for an engaging panel discussion as current users share their unique experiences and talk about the projects they’re standing up within this environment. TBC Tuesday, April 26, 13:30-13:50
OpenStack Talent Development: Lessons Learned What happens when you set out to train hundreds of new OpenStack contributors? Last summer, Intel and Rackspace embarked on this very journey, establishing the OpenStack Innovation Center (OSIC). Designed to drive the enterprise adoption of OpenStack, one of the Center's key tenets is to create a development program aimed to increase the OpenStack talent pool. In this session, we will share our lessons learned while executing on the talent development program. We will walk through our recruitment, training, certification, and mentoring programs, and talk about what worked and what didn't. We will share stories from recent graduates of the program and propose ideas about how others in the community can leverage what we have learned. Michael Apostol (Intel), Tony Campbell (Rackspace) Tuesday, April 26, 14:50-15:30
And Now For Our Regularly Scheduled OpenStack Roadmap Update! OpenStack continues to grow the number of projects in the Big Tent, and each Project is developing at a rapid pace. In this session, we’ll present the OpenStack Product Work Group’s findings across the different projects in an effort to give users a glimpse into OpenStack's upcoming capabilities. The roadmaps presented will span multiple releases and strive to unify the future direction of the individual projects on a per release basis. The findings are collected through direct interviews with the PTLs and condensed into an easy to consume format. You could try to track down what's coming through dozens of etherpads and hundreds of blueprints and spend your days living in IRC. Or you could come to our session. Consider it your ticket to seeing a glimpse into OpenStack's future. Carol Barrett (Intel), Nate Ziemann (IBM), Hugh Blemings (Rackspace) Tuesday, April 26, 15:40-16:20
Cerner’s Journey Towards a Secure Software Defined Data Center Security for Software Defined Data Center is a challenge that a lot of enterprises face. At Cerner we wanted to move to an OpenStack-based SDDC implementation but wanted to ensure that this was done without reducing the security posture of our environment, while at the same time enabling an improved level of operational efficiency and lowered TCO. Some of our key success indicators include the ability to gain increased visibility and control against malicious activity within the Data Center and thus secure East-West traffic, as well as the ability to enable this in an automated manner that allows us to deploy virtual security functions (such as IPS) and easily enforce policies while the virtual security functions continue to maintain an acceptable level of performance. With these top level requirements, we started working on a solution implementation with Intel® Corporation and Midokura® and are in the midst of our engagement. Tarun Viswanathan (Intel), Joe Quint (Cerner Corp.), Giuseppe de Candia (Midokura) Tuesday, April 26, 15:40-16:20
State of the Nation: High Availability for Pets and Hypervisors
By now, high availability for the OpenStack control plane is well understood, and to a large extent is a solved problem which the community continues to refine. In stark contrast, most solutions for compute node HA (for example, where VMs are automatically restarted on a different compute node if there is a failure in the hypervisor or its underlying hardware) are still either relatively immature, experimental, or are in the design phase. This is despite the high demand for this feature, which calls into question the older belief held by some that OpenStack should only accommodate "cattle" VMs which have resilience built in to the application layer.
In this talk, presented by members of the OpenStack HA community who have been collaborating on this topic since Tokyo, we summarize and compare all known current approaches, from vendors including Red Hat, SUSE, Intel, NTT, AWcloud, ChinaMobile, and ZeroStack, and explain the current thinking, challenges presented, and future directions.
Adam Spiers (SuSE), Dawid Deja (Intel) Tuesday, April 26, 15:40-16:20
Turning Pets into Cattle: The Stickiness of Data Many enterprises are interested in migrating their conventional three-tier application (Pets) into a modern cloud-native application (Cattle). At the Tokyo summit, Steve and Leong presented a strategy on how to migrate the web and app-tier of a business-critical application from a visualized world into a hybrid-cloud environment based on OpenStack. In this session, we will continue our discussion and spend some time to talk about data-tier. We will discuss strategies, challenges, and areas that need to be considered when moving applications into the cloud, including private & public clouds. Dr Yih Leong Sun (Intel), Stephen Walli (HP), Gerd Prüßmann (T-Systems) Tuesday, April 26, 15:40-16:20
Upgrading OpenStack?
"Upgrading OpenStack is hard."
"Upgrades take forever."
"You can't skip releases easily."
"Rolling back a failed upgrade is impossible."
These and other myths will be busted in this session, where we'll discuss an upgrade strategy for distributed systems that actually works. Traditional in-place upgrades often entail a mind-bogglingly complex set of steps and ordering that are tough to grok, tricky to automate, and harder to roll back when something goes amiss.  What if there was a workflow that was simple, reliable, easy to revert, and which minimized downtime (while also being less risky for the pager-toting cloud admin)? We'll discuss the blue-green upgrade pattern, a stalwart of distributed systems that we applied to OpenStack (*and it's underpinnings!*) to make upgrades easy.  While we're at it, we'll talk tools, tradeoffs, and feedback from the many clouds we've upgraded successfully in the real world. To top it off, we'll illustrate with a demo of a skip-version upgrade.
Michał Jastrzębski (Intel), Steven Dake, (Cisco), Tom Howley (HP) Eamonn O'Toole (HP), Jan Grant (HPE), Nicholas Randon (HP), Mark Voelker (VMware), Sidharth Surana (VMware) Tuesday, April 26, 16:40-17:20
The Diversity of Innovation This 1-hour panel will focus on the links between diversity and inclusion strategies with technical innovation, talent recruiting and retention, male advocacy, and the OpenStack community. Imad Sousou (Intel), Nithya Ruff (SanDisk), Mark Muehl (Comcast), Dorian Naveh (EMC) Tuesday, April 26, 16:50-17:20
Continuous Compliance & Enforcement of Location-based Policies in an OpenStack Cloud Study after study shows that security is still the biggest barrier to cloud adoption, with compliance and data protection being top concerns for a CISO. Enterprises often have a need to enforce policies (based on data sovereignty requirements) such as virtual workloads should only be placed, or migrated, based on location and assurance of hardware platform. Some others want to launch VMs/Apps/Containers on servers with Boot Integrity—Platform Trust—while some have a need to extend the chain of trust from platform to workload (VM/Container). This session features a customer case study from Visa with technologies from Intel (CIT, TXT and TPM), HPE (Helion OpenStack and Proliant Gen 9 server) and Cloud Raxak (RaxakProtect) that demonstrates end-to-end compliance (boot time to shutdown) using hardware and software attestations. Raghuram Yeluri (Intel), Joy Dorairaj (HPE), Fred Lima (Visa), Rashmi Krishnan (Visa), Prasanna Mulgaonkar (Raxak) Tuesday, April 26, 17:30-18:10
Distributed NFV & OpenStack: Challenges and Potential Solutions At SDN World Congress last year, Peter Willis of British Telecom (BT) presented a set of challenges that are preventing BT from using OpenStack for distributed NFV deployments, specifically the virtual CPE and vBranch use cases. In this session, we will explore the challenges and attempts to identify short- and long-term solutions. Short-term solutions range from creative ways of leveraging Linux capabilities, as well as selective use of OpenStack, OpenDayLight, and other open source projects, while long-term solutions include the identification of promising OpenStack blueprints and projects. Prithiv Mohan (Intel), Adrian Hoban (Intel), Tariq Khan (HP), Peter Willis (British Telecom), Barak Perlman (HPE), Arun Thulasi (HP) Wednesday, April 27, 09:00-09:40
Neutron DSCP: Policing Your Network DSCP is used by large enterprises to manage network traffic intelligently. It provides quality of service to enable traffic engineering, and in some cases, can additionally be relied on to enforce network security policies. The inclusion of DSCP in Newton Neutron now makes this facility available to all users, from individuals to large enterprises. This talk will describe DSCP for those unfamiliar with it, discuss how DSCP is used "in the wild" today, and relate the path taken to deliver the Open vSwitch-based reference implementation for Newton. David Shaughnessy (Intel), Nate Johnston (Comcast) Wednesday, April 27, 09:00-09:40
Building Real-Time Cloud Using OpenStack Users or operators who need to run low-latency or real-time applications, such as audio/video streaming or highly interactive systems, to meet their requirements in clouds might turn to Ironic in OpenStack. However, Ironic lacks the management advantage of virtualization. The good news is that we have sufficient infrastructure to achieve this capability in OpenStack Mitaka, enabling low-latency and real-time applications in OpenStack without the use of Ironic. Nevertheless, it’s still cumbersome to configure and build such real-time clouds. In this session, we share our know-how and experience in building real-time clouds using OpenStack based on our work on real-time KVM and KVM enhancements for NFV at OPNFV. We will show how to configure Nova nodes (CPU, I/O, memory, etc.) and root-cause potential problems with real-time applications on OpenStack. Jun Nakajima (Intel), Yunhong Jiang (Intel) Wednesday, April 27, 09:50-10:30
How to Architect Applications for the Cloud The event logs are usually our best friends for determining whether a system or its component is experiencing problems. With going away from single node setups we have to say goodbye to finding issues with ease just by reading logs here and there. As the number of nodes increase, it becomes painful to track the status of your deployment. Is there anything that might help? Yes, central logging service. With this approach we can quickly get answers to “why” and “where” something went wrong, even when malfunctioning component is offline and there is no way to get to its local storage. But deploying a central logging service might turn out to be a little hassle itself. Luckily, you can take advantage of Docker and Ansible in Kolla, enabling you to have a full centralized logging stack in a matter of seconds. During this presentation we will invite you to the Loggingland and compare different containerized approaches—Logstash and Heka, both paired with Elasticsearch and Kibana. Michał Jastrzębski (Intel), Alicja Kwaśniewska (Intel), Éric Lemoine (Mirantis) Wednesday, April 27, 11:00-11:40
These aren't the logs you're looking for Engage and enable enterprise development teams to realize the value from cloud architecture and accelerate cloud application while enabling solutions that work on any client platform. In this session, we will provide an overview of how to design applications for the cloud and provide lessons learned from a customer who has 800+ applications in various stages of development and deployment. Explore architecture design decisions, key things to consider for integration with existing application, design patterns, code samples, and best practices for developing cloud-aware applications and supporting business strategies for modern application development (cloud, mobile/multi-OS, application security, and improved user experience). William Giard (Intel), Brandon Bohling (Intel) Wednesday, April 27, 11:00-11:40
Is OpenStack Ready for App Developers? The Journey of App Developers in Public Clouds. To make OpenStack more efficient for application developers, the App Ecosystem Working Group studied the current app development experience in OpenStack by conducting cloud benchmarking analysis. Through this analysis, the group compared OpenStack against major cloud providers such as Amazon Web Services, Microsoft Azure, and Google Cloud Platform, seeking to discover best practices, gaps and critical issues in the application development process in OpenStack. Findings are grouped according to the stages of the development life cycle. Marcela Bonell (Intel) Wednesday, April 27, 13:50-14:30
Live Migration, Mitaka and Beyond OpenStack and its functionalities are evolving very fast. Since the Kilo release, Nova has been working on live migration to improve performance, increase stability, implement new features and improve the user experience. To make even more enhancements possible, we have chosen live migration as a priority for the Mitaka cycle. In this session, we will tell you what has changed since the Kilo release in both Nova and QEMU. We will also show how you can configure your OpenStack environment to take advantage of new live migration features and also show common pitfalls you should avoid. And we will give you insight into current work in live migration and what improvements are planned in the future. Pawel Koniszewski (Intel), Paul Murray (HP), Andrea Rosa (HP) Wednesday, April 27, 13:50-14:30
Tired of iptables based security groups? Here's how to gain tremendous speed with Open vSwitch instead! This session will give an overview of current implementations of neutron-openvswitch-agent security group firewall drivers using pure Open vSwitch bridges. We will demonstrate and explain how two approaches work, each using OpenFlow: a security group firewall driver using OpenFlow learn-action rules, and a security group firewall driver using conntrack implemented in Open vSwitch kernel space. We will compare these approachs using benchmarks, explore the pros and cons of each, and talk about tips to debug lost packets in the firewall. Rodolfo Alonso (Intel), Jakub Libosvar (Red Hat) Wednesday, April 27, 13:50-14:30
Building a Flexible OpenStack Cloud From the Ground Up In this session, we will discuss how we build a flexible OpenStack cloud from the ground up to support multiple engineering development workloads at Intel's Open Source Technology Center. We will provide a reference architecture that includes hardware and software components, and will recommend how to deploy an OpenStack Cloud from scratch to meet our flexible compute, storage and networking requirements. We will share the architectural and design decisions that we have made in implementing such a solution. We will provide insight, knowledge, and details of the hardware components that we have chosen for this cloud, how to calculate BOM (bill of materials), operating efficiency (power consumption, battery backup), cloud capacity, and how this architecture can be scaled from an initial 30 nodes cluster to a 100+ nodes cluster. Dr Yih Leong Sun (Intel), Michael Kadera (Intel), John Geier (Intel) Wednesday, April 27, 14:40-15:20
Doubling Performance in Swift with No Code Changes
Would you like to double performance without changing a line of code?
All interpreted languages come to the point in their evolution where a radical turn is needed to optimize performance and a Just In Time (JIT) compilation is the direction they have all turned. We would argue that now is the time for Python. In perhaps an extreme case, Swift’s performance is greatly affected by the latency of Python’s default CPython interpreter. Recently Swift throughput was improved 111% on a 15 node storage cluster and response time was boosted 87%. This was achieved by using PyPy, the JIT version of Python. This talk will describe a proof-of-concept study using PyPy in production with Swift, offer a roadmap for broad adoption of PyPy in the Python world, and provide clear steps for implementing PyPy in Swift.
David Stewart (Intel), John Dickinson (SwiftStack) Wednesday, April 27, 14:40-15:20
OpenStack Strategy: How to Achieve Our Mission
The mission of OpenStack is: "To produce the ubiquitous Open Source Cloud Computing platform that will meet the needs of public and private clouds regardless of size, by being simple to implement and massively scalable."  
Most Stackers work for individual companies with their own OpenStack-powered products and services. What if, for just one hour, we all took off our collective corporate hats (for those who wear them) and focused on what long-term strategies will allow us to achieve OpenStack's mission? No we aren't talking about doing a SWOT analysis. We'll talk about the market, competition and OpenStack's current assets and vulnerabilities. From there, we will begin a brainstorming session to develop potential strategies to guide OpenStack over a multi-year timeframe. All are welcome, whether you are a developer, user, product manager, operator, or business person.
Carol Barrett (Intel), Kenny Johnston (Rackspace) Wednesday, April 27, 15:30-16:10
Architecture and Best Practices to Deploy Hadoop and Spark Clusters with Sahara
Using big data environments as analytics engines to gain useful insight on massive amounts of data in a manageable and timely manner can be a complex challenge. Using Sahara as the orchestration tool helps cloud architects provide high-performance production Hadoop and Spark environments enabling private cloud deployments that are easy for data scientists to manage and consume. We will provide a reference architecture for a Hadoop cluster using Intel-based systems and Sahara focused on performance and ease of use to answer questions regarding security, storage, and supporting future releases without re-installing.
Paul Work (Intel), Sergey Lukjanov (Mirantis), Nikita Konovalov (Mirantis) Wednesday, April 27, 16:30-17:10
Openstack and OpenDaylight: the Current Status and Future Direction One of the biggest OpenDaylight use cases is its integration into OpenStack, which has been the focus of past releases and of the most current release. In this session, we will provide a high-level overview and current status of this integration, as well as new developments and future plans, particularly in relation to the OpenDaylight Beryllium and OpenStack Mitaka releases. The discussion will include support for BGPVPN and L2gateway in OpenDaylight, and networking-odl rewrite in OpenStack, and explore the challenges inherent in coordinating two independent release plans while assuring quality. We will then talk about development and deployment of OpenDaylight with OpenStack, and how to contribute, and wrap up with solicitation of feedback for future plans and development. Isaku Yamahata (Intel), Vishal Thapar (OpenDaylight), Anil Vishnoi (Brocade), Ravindra Kenchappa, Arvind Somya (Cisco), Sam Hague (Red Hat) Wednesday, April 27, 16;30-17:10
Swift Middlewares - What Are They? OpenStack Swift already addresses a wide range of features, and solves challenges that come with storage to a great extent. However, there can be users and operators with new use cases for their specific needs and would wish to have them in Swift. In this scenario, a middleware comes to your rescue and lets you add your own significant feature by extending Swift capabilities. Many features of Swift are implemented as middlewares. It would be worthwhile to understand what they are, how they are implemented and how they communicate with Swift. Mahati Chamarthy (Intel) Thursday, April 28, 09:00-09:40
Magnum or Murano? OpenStack Options for Container Environment Creation and Management
OpenStack users who want to create and manage container environments (like Kubernetes, Docker Swarm, Apache Mesos, etc.) can do so with methods native to each solution’s ecology. They can explore OpenStack Magnum’s API, which enables creation of Kubernetes, Swarm, or Mesos environments, plus abstract management of container workloads. Or they can use Murano, the application catalog, which deploys Kubernetes clusters, pods, and container applications with Wizard ease. Which to choose?
Madhuri Kumari (Intel), Serg Melikyan (Mirantis) Thursday, April 28, 11:00-11:40
Fostering Full Equality Following up on the themes related to gender equality presented at previous summit talks sponsored by the Women of OpenStack Workgroup, this session will focus on our vision of a world where the OpenStack community is truly and completely blind to meaningless differences in gender, color, creed, and culture. What will this new world look like? How will it differ from the current one? A panel of people from all walks of life will explore what that future is and what we can do as a community as a whole to achieve it. Ruchi Bhargava (Intel), Beth Cohen (Verizon), Nithya Ruff (SanDisk), Nina Goradia (IBM), Vince Brunssen (IBM), Josh Kleinpeter (Cisco) Thursday, April 28, 13:30-14:10
Newcomers Need You, How to be a Good Mentor You still remember the first time you heard the word OpenStack, the efforts you had to make to wrap your head around it and to get your first patch merged. Now you know enough to be able to teach other people and bring them up to speed. You would like to be a mentor, but you are not sure about how time consuming it might be, and you still have doubts and questions. In this session, we will try to answer them. We will give advice based on our experience mentoring in the Outreachy and Google Summer of Code programs, suggest some strategies, and list the most common mistakes a new mentor makes. Mahati Chamarthy (Intel), Victoria Martinez de la Cruz (Red Hat), Rossella Sblendido (SuSE) Thursday, April 28, 14:20-15:00
Watcher, a Resource Manager for OpenStack: Plans for the N-release and Beyond Watcher is an open source software package that provides a flexible and scalable resource optimization service for multi-tenant OpenStack-based clouds. Watcher provides a complete optimization loop—including a metrics receiver, optimization processor, and an action plan applier. This provides a robust framework to realize a wide range of cloud optimization goals, including the reduction of data center operating costs, increased system performance via intelligent virtual machine migration, increased energy efficiency, etc. The overall goal is that OpenStack-based clouds equipped with Watcher will decrease their Total Cost of Ownership by way of more efficient use of their infrastructure through targeted optimizations and closed-loop automation. In this talk we will go over the state of Watcher as it is today, its architecture, the team’s accomplishments for the Mitaka release and our plans for the N-release and beyond. Susanne Balle (Intel), Antoine Cabot (bcom), Joe Cropper (IBM), Xavier Priem Thursday, April 28, 14:20-15:00
Dive into Nova Scheduler Performance: Where is the Bottleneck? There are always rumors around Nova scheduler, such as, "The scheduler can handle at most 1k nodes without problems," or, "I don't believe the scheduler can handle many requests using only one tiny process." Some would prefer to launch a couple of scheduler instances to multiply the schedule throughput, but still others might worry about race conditions and decision conflicts that could outweigh the benefits. Cloud users are interested in the deployment strategies to maximize the schedule performance, and developers are curious about the mystery of race condition and its performance impact. Guesses and speculations are everywhere without evidence. So let's do some experiments and let data do the talking! Yingxin Cheng (Intel), Zhenzan Zhou (Intel), Yuntong Jin (Intel), Junwei Liu (China Mobile) Thursday, April 28, 15:10-15:50
DPDK, Collectd & Ceilometer: the Missing Link Between My Telco Cloud and the NFV Infrastructure Collectd exposes statistics that facilitate more resilient and performant telco/NFV clouds, which are vital to monitor systems for malfunctions that could lead to users’ application service disruption, and promptly react to these fault events to facilitate improving overall system performance. By providing Ceilometer with system statistics from collectd, there is more data available to be used for monitoring, performance analysis, fault detection, etc. Using OPNFV Doctor-prescribed enhancements to OpenStack, action can then be taken to negate the effects of any faults in the deployment. Gaps have been identified, and work to improve OpenStack to enable a more fault tolerant cloud environment is well underway. A key part of this work includes expanding the amount of data available about the system (for example, DPDK statistics), and improving alarming functionality in OpenStack Aodh. Emma Foley (Intel), Maryam Tahhan, Carlos Goncalves (NEC) Thursday, April 28, 15:10-15:50
Developing, Deploying, and Consuming L4-7 Network Services in an OpenStack Cloud
RSVP required.
The OpenStack infrastructure services for compute, networking, and storage have matured and are seeing increased adoption in production environments. In this tutorial session we will show you how to combine these fundamental services with other ancillary services around image management, security, monitoring, and policy-based orchestration to build, deploy, and consume rich networking services in your OpenStack Cloud.
This tutorial session is aimed at developers, operators, and users alike. A working environment will be provided to the attendees for experimentation. If you are a developer or vendor, you will learn how to leverage existing OpenStack constructs to expose your service offering. If you are an operator, you will learn how to configure, deploy, and expose these service offerings in a few simple steps. If you are a user, you will learn how to consume single services, or compose a chain of services, and incorporate them in your application topology.
Igor Duarte Cardoso (Intel), Sumit Naiksatam (Cisco), Hemanth Ravi (One Convergence), Ivar Lazzaro (Cisco), Jason Plank (Sungard) Thursday, April 28, 16:10-17:40
Improving Cloud Performance with GCC latest technologies in Clear Linux Project As the Linux community continues to redefine the boundaries of what is possible in a cloud-based Linux distributions running on Intel silicon, both power and performance play an increasingly important role in the industry. In the Clear Linux Project for Intel Architecture, we decided to use the latest GCC technology to boost the performance of a cloud-based system. From profile-based techniques (PGO and AutoFDO) to Function Multi-Versioning (a compiler feature that is capable of optimizing the same code for multiple architectures and then automatically selecting the correct architecture-specific version of the code at runtime) we will show how these technologies can be used to improve runtime performance and unleash the power of IA in Linux distributions. This presentation will help cloud developers to better understand the capabilities of the latest version of GCC compiler. This will include demos and open source templates that can be reused for their own applications. Victor Rodriguez (Intel) Thursday, April 28, 16:10-16:50

Top of page



Quality of Service with Intel® Resource Director Technology 

Quality of Service (QoS) is very important for cloud customers. There are thousands of applications running on the public cloud today where it is really challenging to meet each user's requirements while maintaining efficient use of your environment. Intel introduced several Platform QoS (PQoS) features to monitor and control the shared resource that directly impact tenant services, including Cache Monitoring Technology, Cache Allocation Technology, Memory Bandwidth Monitoring, and more. This demo will show how to use these Intel Architecture features in OpenStack to achieve service objectives. 

Hardware-assisted Secure Containers in OpenStack Clouds

Linux containers/Docker are emerging as a promising complement to virtual machines (VMs) in cloud data centers since they are fast, lightweight, and low-overhead (at runtime and during launch). OpenStack is fast becoming the cloud management software in enterprises and service providers. Enterprises and service providers are looking to launch both VMs and containers using the same control plane. Technologies like Trusted Boot and Trusted Pools based on hardware-root of trust have been mainstream whereby you can have the assurance that VMs/workloads are launching on servers that have demonstrated boot time integrity, essentially addressing a certain set of security concerns. However, for LXC/Docker containers, security continues to be a bottleneck and questions persist. In this live demonstration, learn how containers running in OpenStack clouds use hardware security such as Intel Cloud Integrity Technology and IKGT. This demo will showcase how containers and SDI systems can be aware of and assisted by hardware security features to run better on Intel architecture in OpenStack clouds.

Securing the vEPC Network on OpenStack 

Virtual Evolved Packet Core (vEPC) is an important component of 4G/LTE Telecom networks. Network traffic within the vEPC infrastructure (running on OpenStack) is often in clear, making it pretty easy to maliciously exploit it. Protecting virtual network traffic within vEPC running on an OpenStack installation while leveraging Intel architecture smarts is the goal of this demonstration.

Showcasing VNF Integrity on OpenStack 

Open vSwitch (OVS) and other virtualized network functions are fast becoming one of the backbones of networking within the OpenStack installation, it is imperative that these VNFs are protected from any malicious behavior. Although we can monitor their behavior at runtime, so far there isn’t any way we can be absolutely certain that any given VNF is integrity protected at launch.  Intel’s Cloud Integrity Technology helps us do just that. Not only we can control the whitelisted behavior of VNFs, but we can also guarantee that trusted code, devoid of any changes whatsoever, would be run on the hypervisor whenever the service is started.  

Building Cost-Efficient, High Performance All-flash Block Storage for Your OpenStack Cloud with Intel NVMe SSDs

Network Function Virtualized (NFV) workloads require assured compute and network performance. A solution built with network performance enhanced hardware requires 1. Creating a Network Assurance flavor 2. Openstack conf file and filter scheduler changes 3. Openstack VNF (virtual network function) VM attached to the above flavor which picks the enhanced platform 4. Metrics for network performance at the Neutron, compute, OVS and controller nodes. This demo will showcase orchestration of workload (BNG), performance graphs, and dashboard views after running a workload.

Build towards Diverse Workload Demands with Different Storage Offerings through Virtual Storage Manager

With open source movement blooming, more and more users begin to taste open storage solutions like Ceph. The survey from OpenStack Vancouver summit shows Ceph is the most popular (44%) block storage driver used with OpenStack. Ceph is a scalable, unified storage solution, which expects to help lower TCO. But when people use Ceph in production, they will realize the lack of some kind of management tool, like OpenStack dashboard, can impact daily operation. VSM (Virtual Storage Manager) is a management tool for Ceph developed by Intel, which expects to fill the gap and mitigate the pains to storage administrators. Furthermore, VSM could cooperate with OpenStack to streamline the VM creation with similar operation styles through Web. VSM could also provide storage resources to multiple OpenStack clusters or regions. 

Intel RackScale Architecture Composed FPGA Node

Showcase the innovative Xeon+FPGA platform integrated with Intel Rack Scale Architecture Solution and exposed to the OpenStack environment. The usage scenario is to provision the FPGA accelerated workload in OpenStack. The workload will be able to request the allocation of specific hardware accelerator. SDI layer will expose FPGA acceleration capabilities and cloud systems will be able to pick the free node with acceleration and dynamically reprogram FPGA based on the application demand. OpenStack will use two types of compute nodes: traditional physical servers and logical systems composed in the Intel Rack Scale Architecture rack. The solution is extended by the concept of scalable, secure, robust and independent (in terms of customer privacy exposure) solution for Xeon+FPGA IP (AFU) distribution called the IPStore. Through the IPStore it will be possible to enable four different types of accelerators for workloads run in the cloud.
Automated Fuel Deployment of Intel RackScale Architecture Composed Nodes for DELL G5 Rack

RackScale Architecture is a Intel-led Open architecture for disaggregated pool of servers. Rackscale version 1 is set to launch just before the summit. The most pertinent solution cloud vendors will be looking for is to automate discovery, boot-up, configuration, and deployment of each server node in the rack using standard DMTF interfaces and open source API solutions. More importantly, the benefits of a Rackscale-composed server need to be made available to the OpenStack environment with its unique features so that a user's choice of flavor filters a Rackscale server for their needs and thus provide the best optimized compute, network, and storage solution.

Top of page


Command Presence Workshop Recognized by the Anita Borg Institute and Harvard Business Review, the Command Presence Workshop is intended to teach participants how to handle senior-level, high-pressure meetings. The majority of the workshop is a task force convened to address a simulated, but very realistic crisis event. In the class, each participant assumes a role, and is given about 30 minutes to prepare 1-2 slides. Once the class has prepared, a role-play simulation begins. The workshop participants will experience rapid-fire questioning, interruption, and other situations that occur in senior-level meetings, but in a safe environment conducive to learning techniques for handling such situations. It is not uncommon to hear occasional gasps of disbelief from the participants! Ruchi Bhargava (Intel), Malini Bhandaru (Intel), and a panel of leaders representative of the OpenStack leadership in the community Sunday, April 24, 09:30-12:30
Women of OpenStack Pre-Summit Social Jump start your week at the Summit with a fun evening where you can connect with your fellow female colleagues in the community, hosted by Intel, IBM, and EMC. Join us to learn more about how you can get involved in building a stronger female presence in the OpenStack community. Hosted by Intel, IBM & EMC Sunday, April 24, 18:00-21:00
Fostering Full Equality Session Following up on the themes related to gender equality presented at previous summit talks sponsored by the Women of OpenStack Workgroup, this session will focus on our vision of a world where the OpenStack community is truly and completely blind to meaningless differences in gender, color, creed and culture. What will this new world look like? How will it differ from the current one? A panel of people from all walks of life will explore what that future is and what we can do as a community as a whole to achieve it. Ruchi Bhargava (Intel), Beth Cohen (Verizon), Nithya Ruff (SanDisk), Nina Goradia (IBM), Vince Brunssen (IBM), Josh Kleinpeter (Cisco) Thursday, April 28, 13:30-14:10


Intel Clear Containers: A Breathrough Combination of Speed & Workload isolation Using OpenStack* UEFI Boot Support to Enhance Security in Multi-Tenant Cloud Applications Enterprise-Grade Cloud with Open Source Benefits: How Intel & Red Hat Provide an Enterprise-Grade OpenStack* Solution  The Business Case for Private Cloud: Building a Business Case to Help Ensure the Success of Your Private Cloud Initiative

Top of page


Was this information helpful?