Loading…
This event has ended. View the official site or create your own event → Check it out
This event has ended. Create your own
September 13 -14, 2017 - Los Angeles, CA
Click Here For Information & Registration
View analytic

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

Wednesday, September 13
 

9:00am

Hackathon Registration
Wednesday September 13, 2017 9:00am - 10:00am
Platinum Level - Foyer

10:00am

Hackathon

Join us the day prior to the Keynotes and Breakout Sessions for a full day hackathon.

Day: Wednesday, September 13
Time: 10:00am-4:00pm
Location: Gold 2, JW Marriott LA Live

The focus of this hackathon, while not excluding other ideas, is improving Apache Mesos and it ecosystem.

There will be a number of experienced contributors and committers helping the participants to get started, give direct feedback and if possible even commit code.

Potential areas include:

  • Documentation and website
  • Apache Mesos Core (we will provide a number of tagged beginner issues)
  • Frameworks and DC/OS SDK
  • Integration (e.g., Cloud Provider templated)
  • Apache Mesos UI
  • CLI
  • Benchmarking
  • Monitoring
  • Tooling (e.g., local vagrant or docker setups)

All other proposals are welcome as well and will be discussed at the beginning of the hackathon.


Wednesday September 13, 2017 10:00am - 4:00pm
Gold Level - Salon 2

3:00pm

Pre-Registration
Wednesday September 13, 2017 3:00pm - 5:00pm
Platinum Level - Foyer
 
Thursday, September 14
 

8:00am

Continental Breakfast
Thursday September 14, 2017 8:00am - 9:00am
Diamond Level - Foyer

8:00am

Registration
Thursday September 14, 2017 8:00am - 5:30pm
Platinum Level - Foyer

9:00am

9:10am

Keynote Panel Discussion: Mesos in the Enterprise - Moderated by Josh Bernstein, VP Technology, {code}

Come hear how Apache Mesos is influencing the IT infrastructure at large enterprises, panelists include:

  • Michael Aguiling (CTO at JPMC)

  • Larry Rau (Director of Architecture at Verizon)

  • Cathy Daw (Director of Engineering at Mesosphere)

  • Stefan Bauer (Head of Development Data Analytics at Audi)

  • Hubert Fisher (Audi)

  • Moderated by Josh Bernstein (VP Technology at {code})


Speakers
avatar for Josh Bernstein

Josh Bernstein

VP of Technology, {code}
Josh is an open source advocate and lifelong technologist. As the VP of Technology for Dell, he’s at the helm of {code}, the open source arm of the organization focused on advancing emerging technologies to support software-based infrastructures.  Prior to Dell, Josh ran the Siri... Read More →
avatar for Hubert Fischer

Hubert Fischer

Senior IT Project Manager, Audi Electronics Venture GmbH


Thursday September 14, 2017 9:10am - 9:40am
Diamond Salon 4 & 5

9:40am

Keynote: One Platform for the Containerized Future - Tobi Knaup, CTO and Co-Founder, Mesosphere
Speakers
TK

Tobi Knaup

Mesosphere
Tobi Knaup is a co-founder of Mesosphere, a startup that aims to simplify the development and operation of large scale infrastructure and distributed systems. He is the main author of Marathon, an Apache Mesos framework for managing long-running services. Tobi was one of the firs... Read More →


Thursday September 14, 2017 9:40am - 10:00am
Diamond Salon 4 & 5

10:00am

Keynote Panel Discussion: SMACK Today and Tomorrow - Moderated by Ben Hindman, Co-Creator, Apache Mesos and Founder, Mesosphere

The SMACK stack is becoming the architecture for fast data processing and we’re bringing together some of the creators of the projects to talk about how SMACK is impacting the data analytics landscape.

  • Neha Narkhede (co-founder and CTO of Confluent)

  • Jonathan Ellis (co-founder and CTO of Datastax)

  • Benjamin Hindman (co-founder at Mesosphere)


Speakers
avatar for Jonathan Ellis

Jonathan Ellis

CTO, DataStax
Jonathan Ellis is the founder and CTO of DataStax , and was Project Chair of Apache Cassandra for over six years. Before his work on distributed databases, he built a multi-petabyte object store based on Reed-Solomon encoding for backup provider Mozy.


Thursday September 14, 2017 10:00am - 10:30am
Diamond Salon 4 & 5

10:30am

11:00am

Lessons Learned: Deploying Microservices Software Product in Customer Environments - Mark Galpin, JFrog, Inc.
Any team with strong DevOps practices wants to move to a microservices model to get the most out of them. But honestly, a lot of DevOps, and the best practices laid out for microservices, assumes that you have control over the deployment environment. But what if you don’t?
At JFrog we developed a microservices, cloud-native container-based design, and then we had to help our customers deploy it to diverse environments. Some of which didn’t allow containers. We will discuss our love for Apache Mesos and DC/OS, and why its always central to our thinking in deployment design, but what we do to help customers who can’t use it. What did we learn? What are we taking into older products we are now trying to break into microservices? What do we take forward? What mistakes have we sworn never to make again? Come and find out!

Speakers
avatar for Mark Galpin

Mark Galpin

Senior Solution Engineer, JFrog, Inc.
Mark Galpin is a Senior Solution Engineer at JFrog, experts in binary management and distribution making the products JFrog Artifactory, JFrog Bintray, and JFrog Xray. Before coming to JFrog in 2015, he spent eight years working on software, interoperability, and data standards a... Read More →


Thursday September 14, 2017 11:00am - 11:40am
Diamond Salon 6

11:00am

IP Anycast: DC/OS Networking for the Modern Datacenter - Bill Green, New Relic, Inc.
Modern datacenter networks are leveraging BGP and spine-leaf topologies to enable massive scale-out. This presentation will survey New Relic's implementation of IP Anycast for Docker containers running on DC/OS. Such an implementation allows several containers to be deployed with the same IP address, allowing the network to share traffic between them. We will explore some container deployments that are greatly simplified by using this technique.

Speakers
avatar for Bill Green

Bill Green

Senior Site Reliability Engineer, New Relic, Inc.


Thursday September 14, 2017 11:00am - 11:40am
Diamond Salon 3

11:00am

Applying Geospatial Analytics at a Massive Scale using Kafka, Spark and Elasticsearch on DC/OS - Adam Mollenkopf, Esri
This session will explore how DC/OS and Mesos are being used at Esri to establish a foundational operating environment to enable the consumption of high velocity IoT data using Apache Kafka, streaming analytics using Apache Spark, high-volume storage and querying of spatiotemporal data using Elasticsearch, and recurring batch analytics using Apache Spark & Metronome. Additionally, Esri will share their experience in making their application for DC/OS portable so that it can easily be deployed amongst public cloud providers (Microsoft Azure, Amazon EC2), private cloud providers and on-premise environments. Demonstrations will be performed throughout the presentation to cement these concepts for the attendees.

Speakers
avatar for Adam Mollenkopf

Adam Mollenkopf

Real-Time & Big Data GIS Capability Lead, Esri
Adam Mollenkopf is responsible for the strategic direction Esri takes towards enabling real-time and big data capabilities in the ArcGIS platform. This includes having the ability to ingest real-time data streams from a wide variety of sources, performing continuous and recurring... Read More →


Thursday September 14, 2017 11:00am - 11:40am
Diamond Salon 1 & 2

11:00am

My Application is Ready for Production... Now What? - Gastón Kleiman, Johannes Unterstein, Kevin Klues, & Philip Norman, Mesosphere (limited spots, pre-registration suggested)
We'll explore the options offered by Mesos, Marathon, and DC/OS to try to answer all these questions and more:
  • How do I deploy my application on a Mesos cluster? 
  • How can I monitor my application and make sure that everything stays running happily?
  • How can I monitor the underlying infrastructure? 
  • Help! Something strange is going on with my tasks! How do I debug them? 
  • Uhm, my tasks are now spread across my entire cluster… how do I deal with all the logs?
  • Wait, how do my services know how to talk to one another? 
  • How can I prevent downtime during deployments?

Please indicate your interest by completing this interest form

For all University sessions, attendees will be asked to bring a laptop with the ability to ssh into a Linux machine.

Speakers
avatar for Gastón Kleiman

Gastón Kleiman

Distributed Systems Engineer, Mesosphere
Gastón Kleiman is a Senior Distributed Software Engineer at Mesosphere - he used to work on Marathon, but is focusing on Apache Mesos these days. Gastón fell in love with distributed systems while contracting for Google and working with Borg, MapReduce and other cool technology... Read More →
avatar for Kevin Klues

Kevin Klues

Tech Lead Manager, Mesosphere
Kevin Klues is a Tech Lead Manager at Mesosphere running the DC/OS ClusterOps team. Since joining Mesosphere, Kevin has been involved in the design and implementation of a number of Mesos’s core subsystems, including GPU isolation, Pods, the Mesos CLI and Attach/Exec support. H... Read More →
avatar for Philip Norman

Philip Norman

Software Engineer, Mesosphere
Philip runs the DC/OS day 2 operations working group, and is responsible for the DC/OS metrics project. Philip says: "Come talk to me about monitoring containers!"
avatar for Johannes Unterstein

Johannes Unterstein

Distributed Applications Engineer, Mesosphere
Johannes Unterstein is a Distributed Applications Engineer at Mesosphere in Hamburg where he works on Mesosphere's Marathon project.



Thursday September 14, 2017 11:00am - 12:30pm
Diamond Salon 7

11:50am

Adopting Microservices in Healthcare, Breaking Apart the Monolith - Aaron Baer, athenahealth
Healthcare systems traditionally lag behind from the curve of technology trends. Information is plentiful in the space of healthcare and data is typically in large silos. In this presentation we'll explore how our DevOps team is adopting microservices, breaking apart it's monolith, and adopting a new data fabric to allow for massive scale and increased innovation in the space of healthcare information as part of it's platform infrastructure platform. We'll talk about the architecture and tools chosen as well as what went well and what didn't along with some of the efforts still on going as part of this change.

Speakers
avatar for Aaron Baer

Aaron Baer

Lead Member of Technical Staff, athenahealth, Inc.



Thursday September 14, 2017 11:50am - 12:30pm
Diamond Salon 6

11:50am

Day 2 Operations Best Practices with Apache Mesos - Janet Yu, SignalFx & Ben Lin, Mesosphere
You’ve got a Mesos cluster running, now what? Keeping up with a dynamic production environment requires an arsenal of flexible tools. Standing up the cluster is just the beginning. How do you know that the cluster is healthy and behaving as expected? What metrics should be measured to proactively discover inefficiencies and bottlenecks in the system?

This is the realm of Day 2 Operations, the critical piece that keeps your applications humming and your pager silent. Common challenges that are faced in Day 2 Operations will be discussed, with a demo highlighting a fully integrated solution.

Speakers
BL

Ben Lin

Solutions Architect, Mesosphere
Ben Lin is the APAC Tech Lead for Mesosphere. Ben was one of the first Solutions Architects at Mesosphere and worked closely with customers on architecture design and enablement. He standardized and built out the Mesosphere consulting services portfolio. Prior to Mesosphere, Ben... Read More →
avatar for Janet Yu

Janet Yu

Software Engineer, SignalFx
Janet Yu works on the UI and REST API layers at SignalFx. Last year, she gave a lightning talk at MesosCon North America. Before that, she contributed to the OpenStack dashboard, networking, and cloud policy projects while at VMware. And before that, she was in the Junos Manageab... Read More →



Thursday September 14, 2017 11:50am - 12:30pm
Diamond Salon 3

11:50am

State is Hard: An SDK for Building Stateful Applications - Gabriel Hartmann, Mesosphere, Inc.
Apache Mesos and DC/OS are powerful tools to manage, deploy, and maintain services. But, rolling your own stateful application on top of DC/OS requires a deep understanding of Apache Mesos primitives and DC/OS components. Enter the DC/OS SDK.

From a bird’s eye view, most stateful systems look quite similar. Kafka, Cassandra, HDFS, Elastic and the rest of Big Data systems all have their own very specific concerns, but fundamentally they all need to:

The DC/OS SDK codifies the 95% of code that is shared between most services, standardizing how those services interact with Mesos.

In this talk, we will present an overview of the interface and briefly demonstrate how to write a service. We'll also present deep dive on the the internal design and implementation of services.

Speakers
GH

Gabriel Hartmann

Technical Lead, Mesosphere, Inc.
Gabriel is the lead engineer on the DC/OS Stateful SDK team at Mesosphere. Prior to Mesosphere, Gabriel worked on SQLAzure, the best database in the world. Before Mesosphere he worked at Microsoft working on the high availability team developing the second generation of SQL Azure... Read More →


Thursday September 14, 2017 11:50am - 12:30pm
Diamond Salon 1 & 2

12:30pm

Lunch (Attendees on Own)
Thursday September 14, 2017 12:30pm - 2:00pm
TBA

2:00pm

Deployment API - Chris Mays & Micah Noland, HERE Technology
HERE Technologies has many teams deploying services to DC/OS. When first using Mesos, and now DC/OS, the ops team owned the deployments but this caused bottlenecks. Our goal then became to allow the dev teams to deploy directly to DC/OS using an API. However, the ops team didn’t feel the dev team should be exposed to the full Marathon API. First, there would be a learning curve, both now and in the future when new features are added. Second, operationally there are certain things that NEED to be done correctly. For example, certain arguments need to be passed to the docker daemon for logging and monitoring to be setup correctly in the environment. For these reasons, the Deployment API was created to sit between the developers and DC/OS. This allows the operations team to “enhance” the Marathon JSON and provide standards across many development teams.

Speakers
CM

Chris Mays

Principal Engineer, HERE Technologies
Chris Mays is a Principal Engineer at HERE Technologies. He has worked at here for 12 years writing automation, architecting solutions and playing with Docker and Mesos.


Thursday September 14, 2017 2:00pm - 2:40pm
Diamond Salon 6

2:00pm

GASP! My Container Cluster Just Went Down and I Don’t Know Why! - Kamalakannan Muralidharan, PayPal & Ranga Rajagopalan, Avi Networks
Traffic management in a container environment, especially for microservices involves more than just load balancing. Load balancing traffic from client to core service (north-south) along with inter-service and inter-cluster interactions (east-west) is key.

There are different ways of load balancing methods in a container architecture:
North-south (Fronting LB)
East-West (Client side)

The primary challenges with load balancing in a dynamic environment (like cluster manager environment) are the consequences of a container outage.
How do I do IP per POD vs Port Mapping?
How can I enforce dynamic registration as part of the container lifecycle?

Speakers
KM

Kamalakannan Muralidharan

Head of Product Marketing, PayPal
Kamalakannan Muralidharan, Sr. Member of Technical Staff, PayPal | I have 13 years of experience in Software Engineering. We have been using Mesos and its framework in our company for the past 3 years. I have presented about OSGi in JavaOne and also about Docker and distributed... Read More →
RR

Ranga Rajagopalan

CTO, Avi Networks
Ranga Rajagopalan, CTO, Avi Networks | Over the last 15 years prior to co-founding Avi Networks, Ranga has been an architect and developer of several high-performance distributed operating systems as well as networking and storage data center products. Before his current role as... Read More →



Thursday September 14, 2017 2:00pm - 2:40pm
Diamond Salon 3

2:00pm

Deploying Kafka on DC/OS - Kaufman Ng, Confluent
Apache Kafka is increasingly popular as the streaming platform of choice for real-time data pipelines. In addition, Kafka and microservices are deployed together in DC/OS. In this presentation, Kaufman Ng will discuss the best practices on deploying Kafka to DC/OS, the challenges, and lessons learned from customer deployments.

Speakers
avatar for Kaufman Ng

Kaufman Ng

Solutions Architect, Confluent
Solutions Architect at Confluent. Contributor to Apache Kafka.



Thursday September 14, 2017 2:00pm - 2:40pm
Diamond Salon 1 & 2

2:00pm

Bootstrapping Secure Mesos Clusters - Vishnu Mohan, Adam Bordelon, Vinod Kone & Jorg Schad, Mesosphere (limited spots, pre-registration suggested)
Apache Mesos offers a range of features that allow you to secure your cluster and prevent breaches and other attacks. This talk will provide an overview of the security features and recommendations for hardening your Mesos cluster.

The goals of for this talk are to depict how to:
1. Isolate the cluster perimeter with strong authentication and authorization across all interfaces.
2. Secure and protect the internal cluster communication, containers, and sandboxes.
3. Enhance cluster security with support for 3rd party security integrations.

Topics to cover include:
1. Security Zones
2. Security Modes
3. User and Framework/Service Authentication
4. User and Framework/Service Authorization
5. TLS Encryption
6. Namespaces
7. Secrets Management
8. Hardening Recommendations


Please indicate your interest by completing this interest form.

For all University sessions, attendees will be asked to bring a laptop with the ability to ssh into a Linux machine.


Speakers
AB

Adam Bordelon

Mesosphere, Distributed System Engineer
Adam Bordelon is a distributed systems architect at Mesosphere and an Apache Mesos committer. Before joining Mesosphere, Adam lead development on Hadoop core at MapR, built distributed systems for recommendations at Amazon, and re-architected the LabVIEW compiler at National Inst... Read More →
avatar for Vinod Kone

Vinod Kone

Tech Lead and Engineering Manager, Mesosphere, Inc.
Vinod Kone is a committer and PMC member of the Apache Mesos project. He is currently a Tech Lead and Engineering Manager @ Mesosphere. Previously, he was a Tech Lead and Manager of the Mesos team @Twitter. Vinod completed his PhD in Computer Science from UC Santa Barbara.
VM

Vishnu Mohan

Solutions Architect, Mesosphere
Vishnu Mohan is a Solutions Engineer at Mesosphere working closely with strategic customers in the field on Performance, Management and Automation initiatives. When he’s not analyzing sandbox logs or coding against some API for fun, he may be found rock climbing, hiking, runnin... Read More →
avatar for Jörg Schad

Jörg Schad

Software Engineer, Mesosphere
Jörg is a software engineer at Mesosphere in Hamburg. In his previous life he implemented distributed and in memory databases and conducted research in the Hadoop and Cloud area. His speaking experience includes various Meetups, international conferences, and lecture halls.



Thursday September 14, 2017 2:00pm - 3:30pm
Diamond Salon 7

2:50pm

API Gateway Pattern & Kong in a Microservices World - Marco Palladino, Mashape
Kong (https://getkong.org/) is the most widely adopted OSS gateway for APIs and
Microservices. Built on top of Lua and NGINX, Kong can is a high-performance
gateway that can be elastically deployed behind the firewall to secure, protect
and extend RESTful APIs and Microservices via Kong Plugins.

In a container world APIs are becoming increasingly more important as a
communication medium - inside and outside the firewall. The more services are
being created, the harder it gets to efficiently secure, manage and extend them
in a variety of environments, in singe or multi-DC setups.

API gateways can be used to centralized common functionality in one place, by
providing a highly performant, extensible and lightweight layer for both
internal and external services, including serverless functions.

Speakers
avatar for Marco Palladino

Marco Palladino

CTO, Kong
Marco Palladino is an inventor, software developer, and internet entrepreneur. He is the co-founder and CTO of Mashape, which was started in 2010 and is based in San Francisco, California. He is currently responsible for the design and delivery of the Mashape products, while also... Read More →



Thursday September 14, 2017 2:50pm - 3:30pm
Diamond Salon 6

2:50pm

Container Networking for Micro-Services: An Apache Mesos and DC/OS Networking Deep Dive - Deepak Goel & Jörg Schad, Mesosphere
Apache Mesos and DC/OS allow users to deploy distributed applications and in particular micro-services across a large cluster. Therefore, networking becomes an important aspect especially when trying to provide highly-available applications on top of an unreliable infrastructure.

In this talk we will first present the various challenges around networking for distributed micro-service architectures, including

* Connectivity
* Service Discovery
* Load-balancing
* Isolation

Next, we will discuss different solutions to those problems with the example of DC/OS including:
* Minuteman
* MesosDNS
* Marathon-LB
* Overlay network

As for most of the above challenges there is not a one-size-fits-all solution we have an in-depth look at the trade-offs between different solutions.

Speakers
DG

Deepak Goel

Software Engineer, Mesosphere Inc.
avatar for Jörg Schad

Jörg Schad

Software Engineer, Mesosphere
Jörg is a software engineer at Mesosphere in Hamburg. In his previous life he implemented distributed and in memory databases and conducted research in the Hadoop and Cloud area. His speaking experience includes various Meetups, international conferences, and lecture halls.



Thursday September 14, 2017 2:50pm - 3:30pm
Diamond Salon 3

2:50pm

From Shore to Ship: Using Mesosphere Enterprise DC/OS to Deliver Real-Time Microservices to a Global Fleet of Ships - Eli Tsinovoi, EY
Since 1968, the Royal Caribbean, Celebrity Cruises, and Azamara cruiseline brands have redefined vacationing, guest experience, and innovation. Each successive class of ships is an architectural marvel and pushes the envelope of what is possible. To capture the imagination of families and adventurous vacationers, this global cruise line debuted countless industry “firsts” – such as rock climbing, ice skating and surfing at sea. Royal Caribbean isn’t just at the forefront of vacation innovation; it is also driving a tech revolution

This presentation will describe why Royal Caribbean chose Mesosphere Enterprise DC/OS as a their platform for delivering a seamless guest experience through deployment of distributed microservices, leveraging Reactive Architecture built using Lagom, Cassandra and Kafka. We will explain how the simplicity and intuitiveness of the DC/OS platform has help

Speakers
avatar for Eli Tsinovoi

Eli Tsinovoi

Microservices & Data Streaming Solution Leader, EY
Eli Tsinovoi is a Microservices Solution Leader who has been partnering with Royal Caribbean Cruises Ltd., a global cruise company based in Miami, Florida. As one of the world’s largest cruise line operators, Royal Caribbean strives to offer first class service to all cruise gu... Read More →


Thursday September 14, 2017 2:50pm - 3:30pm
Diamond Salon 1 & 2

3:30pm

Afternoon Coffee Break
Thursday September 14, 2017 3:30pm - 4:00pm
Diamond Level - Foyer

4:00pm

Continuous Delivery for DC/OS with Spinnaker - Will Gorman, Cerner
In this presentation, Will Gorman will provide an overview of the Spinnaker continuous delivery platform, and discuss how extending it to support DC/OS as a deployment target is helping Cerner to manage the challenges of deploying its healthcare services architecture safely and reliably.

Speakers
avatar for Will Gorman

Will Gorman

Software Architect, Cerner
Will Gorman is a software architect at Cerner where he has been building health care solutions for over a decade. More recently he has been working on transforming the deployment of those solutions with DC/OS. He loves to create tools to help other developers to build and deploy... Read More →


Thursday September 14, 2017 4:00pm - 4:40pm
Diamond Salon 6

4:00pm

Apache Mesos and Security: The Joy of Mesos with Calico - Diego Oliveira & Acácio Alves dos Santos, PagSeguro
Taking care of security is key point, even more when working with someone else's money. PagSeguro (http://pagseguro.uol.com.br/) takes that to the next level, putting an additional security policy in it's Apache Mesos cluster with project Calico (https://www.projectcalico.org/). In this presentation it shows how to secure it's services, integrate calico routing propagation (BGP) with an on premises legacy network infrastructure, link with cloud providers (AWS, Azure) and shows the gains in time to release new ACLs compared with the old model.

Speakers
avatar for Diego Olievira

Diego Olievira

PagSeguro
Diego Oliveira is a senior technical manager at PagSeguro, a payment service provider in Brazil under UOL Inc. holding. He is a constant speaker at local conferences like TDC and QConSP. In the current position Diego is the principal engineer working with a team of more then 300... Read More →
avatar for Acácio Santos

Acácio Santos

PagSeguro
Acácio Santos is the SRE Coordinator at PagSeguro, a payment service provider in Brazil under UOL Inc. holding. He is passionate about DevOps culture and a "keep it simple" minded person. His team is responsible for maintaining the product's PaaS and DevOps tools.



Thursday September 14, 2017 4:00pm - 4:40pm
Diamond Salon 3

4:00pm

DataStax Enterprise on DC/OS - Yes, It’s Possible; Customer Case Studies - Ravi Yadav, Mesosphere & Kathryn Erickson, DataStax
Migrating a database to a containerized infrastructure brings with it a whole host of challenging new issues. Concerns range from persistence management, availability requirements, and complicated recovery semantics. If you’re considering migrating a database to containers in the future, this talk should enlighten the path.

We will walk you through the current state of the DataStax Enterprise framework. We'll explore the details of the integration between DSE and DC/OS and cover tradeoffs between different deployment options. We’ll tell the real story of the evolution of this framework. In doing so we’ll discuss how our initial design decisions conflicted with customer expectations and how we worked through these and other engineering feats to get to the offering we have today. We'll then tell a few customer deployment stories and conclude with a demo of DSE on DC/OS.

Speakers
avatar for Kathryn Erickson

Kathryn Erickson

Datastax
Kathryn Erickson leads the Technology Partnership Program for DataStax. Prior to joining DataStax, Kathryn was a leader within the Sales Engineering team at Fusion-io and prior to that she conducted and patented research at the MITRE Corporation. She also worked as part of the Cy... Read More →
avatar for Ravi Yadav

Ravi Yadav

Platform Partnerships Lead, Mesosphere, Inc.
Ravi Yadav is responsible for platform partnerships and ecosystem development at Mesosphere. Prior to this, he was a Developer Advocate at IBM and worked on IBM Watson partnerships and advocacy. In his past life, he worked on developing drivers for medical devices.


Thursday September 14, 2017 4:00pm - 4:40pm
Diamond Salon 1 & 2

4:00pm

Building Your First Stateful DC/OS Service - Ben Wood, Mesosphere (limited spots, pre-registration suggested)
Apache Mesos and DC/OS are powerful orchestration tools. But, building stateful application on top of DC/OS requires understanding Apache Mesos primitives and DC/OS components, and writing complex scheduler code.

From a bird’s eye view, most stateful systems look similar. Apache Kafka, Apache Cassandra, and other stateful systems have their own concerns, but they all need to provision storage, scale capacity, be discoverable by clients, be manageable by operators, and be resilient to failure (which is complicated for databases).

In this talk, Ben will demonstrate by example how the DC/OS SDK enables you to build stateful applications which satisfy these common requirements. Particular attention will be paid to how the abstractions of the DC/OS SDK free developers from focusing on interfacing with Apache Mesos and DC/OS and instead concentrate on building a robust stateful service.

Please indicate your interest by completing this interest form.

For all University sessions, attendees will be asked to bring a laptop with the ability to ssh into a Linux machine.

Speakers
BW

Ben Wood

Ben is a software engineer on the DC/OS SDK team at Mesosphere. Prior to Mesosphere, Ben was a software and infrastructure engineer at Autodesk. He has also been an engineer and technical owner of distributed performance measurement systems at Neumob and SOASTA.


Thursday September 14, 2017 4:00pm - 5:30pm
Diamond Salon 7

4:50pm

Apache Mesos For All Your Hybrid Cloud Needs - Imran Shaikh, YP
Hybrid cloud is bigger than ever. The demand is skyrocketing especially after the outage of a leading public cloud provider earlier in the year. It served us a reminder that moving to the cloud does not absolve us from the responsibility of business continuity and disaster recovery planning of our apps. Unfortunately, none of the solutions that are out there lets you run a hybrid cloud setup with ease. They are still: first generation, difficult to install, operate or administer. What’s needed is an extremely flexible and a highly scalable model to run modern applications across on-premise data centers, private or public clouds. Lots have change with the advent of containers, serverless architectures, message queues, IoT, CI/CD and big data jobs. At YP, we have leveraged Mesos to run all kinds of workload with ease while retaining full flexibility and better business continuity planning.

Speakers
avatar for Imran Shaikh

Imran Shaikh

Lead/Architect
Imran is a Lead/Architect working in Greater LA area. He is a proven technical leader with an industry experience of 14+ years working with Fortune 500 companies. He has worked extensively developing, architecting and managing cloud and distributed computing technologies at Yahoo... Read More →


Thursday September 14, 2017 4:50pm - 5:30pm
Diamond Salon 6

4:50pm

What Security People Want: Making DevSecOps Happen with Containers - Tsvi Korren, Aqua Security
Any project of moving an application to run in containers will eventually run into the security wall. Why is it so hard to get security on board? This presentation will discuss why containers are a disruption for security, and how to address their concerns in a containerized environment. Learn how to have a conversation with your security people, and together embrace the opportunity to execute better security through automation and collaboration.

Speakers
avatar for Tsvi Korren

Tsvi Korren

Director, Solution Architecture, Aqua Security
Tsvi Korren, CISSP, has been an IT security professional for 20 years with background in business process consulting in large organizations. Most recently at CA Inc., he worked across verticals in government, retail, financial institutions and healthcare to implement compliance a... Read More →


Thursday September 14, 2017 4:50pm - 5:30pm
Diamond Salon 3

4:50pm

Building Next-Generation Audience Targeting and Analytics at NBCUniversal - Thomas Barr, NBCUniversal
NBCUniversal is swimming in data. From viewership numbers to behavior, the media juggernaut depends on accurate, real-time data analytics to sell advertisers on its ability to target audiences with precision. As Chief Architect of Audience Studio, Thomas Barr will discuss how he built a data analytics cloud leveraging the application-aware scheduling of Apache Mesos and DC/OS to deliver the next-generation of personalized, predictive data services. Using technologies like Spark, Kafka, and Cassandra, the Audience Studio team empowers data scientists to build targeted audience profiles with myriad data sets, both batch and streaming, in real-time.

Speakers
TB

Thomas Barr

NBCUniversal



Thursday September 14, 2017 4:50pm - 5:30pm
Diamond Salon 1 & 2

5:30pm

7:00pm

Town Hall: Apache Mesos (includes Networking and Storage topics) - Jie Yu, Mesosphere
What’s a community gathering without an opportunity for the community to gather and discuss pressing needs and issues? This year MesosCon will offer town halls for Mesos and Marathon/Chronos -- where folks from the community can connect with each other face-to-face and help shape the future direction of the projects. Town Halls will begin at 7:00pm after the onsite reception and will include drinks and appetizers sponsored by Mesosphere. 

Speakers
avatar for Jie Yu

Jie Yu

Tech Lead, Mesosphere
Jie Yu is a Tech Lead at Mesosphere, Inc, focused on containerization, storage and networking. Before joining Mesosphere, he was a software engineer at Twitter. Jie obtained his PhD in Computer Science and Engineering from the University of Michigan where he conducted research fo... Read More →


Thursday September 14, 2017 7:00pm - 8:30pm
Diamond Salon 8 & 9

7:00pm

Town Hall: DC/OS - Judith Malnick, Mesosphere
What’s a community gathering without an opportunity for the community to gather and discuss pressing needs and issues? This year MesosCon will offer town halls for Mesos and Marathon/Chronos -- where folks from the community can connect with each other face-to-face and help shape the future direction of the projects. Town Halls will begin at 7:00pm after the onsite reception and will include drinks and appetizers sponsored by Mesosphere. 

Speakers

Thursday September 14, 2017 7:00pm - 8:30pm
Diamond Salon 7

7:00pm

Town Hall: Marathon/Chronos - Johannes Unterstein, Mesosphere
What’s a community gathering without an opportunity for the community to gather and discuss pressing needs and issues? This year MesosCon will offer town halls for Mesos and Marathon/Chronos -- where folks from the community can connect with each other face-to-face and help shape the future direction of the projects. Town Halls will begin at 7:00pm after the onsite reception and will include drinks and appetizers sponsored by Mesosphere. 

Speakers
avatar for Johannes Unterstein

Johannes Unterstein

Distributed Applications Engineer, Mesosphere
Johannes Unterstein is a Distributed Applications Engineer at Mesosphere in Hamburg where he works on Mesosphere's Marathon project.


Thursday September 14, 2017 7:00pm - 8:30pm
Diamond Salon 6
 
Friday, September 15
 

8:00am

Continental Breakfast
Friday September 15, 2017 8:00am - 9:00am
Diamond Level - Foyer

8:00am

Registration
Friday September 15, 2017 8:00am - 4:30pm
Platinum Level - Foyer

9:00am

Keynote: Welcome to Day 2
Friday September 15, 2017 9:00am - 9:10am
Diamond Salon 4 & 5

9:10am

Keynote Panel: Future of Cluster Management - Moderated by Elizabeth Joseph, Open Source Developer Advocate, Mesosphere

We’ve assembled representatives from organizations operating some of the largest Mesos clusters we know to talk about where they think the future of cluster management is going. You’ll hear from:

  • Sharma Podila (Distributed Systems Software Architect at Netflix)

  • Sam Eaton (VP of Engineering, Operations, and Infrastructure at Yelp)

  • Ian Downes (Sr. Engineering Manager at Twitter)

  • Zhitao Li (Senior Software Engineer at Uber)

  • Moderated by Elizabeth Joseph (Open Source Developer Advocate at Mesosphere)


Speakers
avatar for Ian Downes

Ian Downes

Senior Engineering Manager, Twitter
avatar for Sam Eaton

Sam Eaton

VP of Engineering - Operations & Infrastructure, Yelp
avatar for Elizabeth K. Joseph

Elizabeth K. Joseph

Developer Advocate, Mesosphere
Elizabeth K. Joseph is a Developer Advocate at Mesosphere focused on DC/OS and Apache Mesos, and runs OpenSourceInfra.org. Previously, she worked for a decade as a Linux Systems Administrator, spending the past four years working at HPE on the OpenStack Infrastructure team. She i... Read More →
ZL

Zhitao Li

Senior Software Engineer, Uber
Zhitao Li is a senior software engineer at Uber Inc. In his two years career, he was both member of the team containerizing the service oriented architecture with Docker, and also among the initial effort of adopting Mesos to modernize the cluster management stack. Before Uber, Z... Read More →
avatar for Sharma Podila

Sharma Podila

Senior Software Engineer, Netflix
Sharma Podila is a Senior Software Enginner in the Edge Engineering team at Netflix, Inc. His current work includes Fenzo, an open source generic scheduler with plugin based optimizations, developing cloud native Mesos frameworks, and evolving the microservices platform at Netfli... Read More →


Friday September 15, 2017 9:10am - 9:40am
Diamond Salon 4 & 5

9:40am

Keynote: NSG Data Analytic Architecture Service - Todd Myers & Kevin Fitzhenry, National Geospatial-Intelligence Agency
In this presentation, Todd Myers and Kevin Fitzhenry will present how the National Geospatial-Intelligence Agency is adopting mesos for large scale processing.

Speakers
avatar for Kevin Fitzhenry

Kevin Fitzhenry

National Geospatial-Intelligence Agency
System Engineer contractor that lead the cloud transition team for National Geospatial-Intelligence Agency (NGA) Research division. Now responsible for implementing production DC/OS clusters.
avatar for Todd Myers

Todd Myers

Senior Architect, National Geospatial-Intelligence Agency
Senior Architect leading the deployment of our multi datacenter Mesos processing clusters for the National Geospatial-Intelligence Agency.



Friday September 15, 2017 9:40am - 10:00am
Diamond Salon 4 & 5

10:00am

Keynote: Are We There Yet? - Ross Gardler, Executive VP, ASF and Principal Program Manager, Azure Container Service, Microsoft

Orchestration of workloads is a hot topic. Everywhere we look there are opinions. All orchestration options claim to be the best, all options have their limitation. We listen to conference presentations that tell us how Company Foo has had amazing results with solution X and that they would never use solution Y, but the very next session tells us that equally impressive results have been achieved with solution Y and they'd never use X. What are we supposed to choose? Who are we supposed to listen to? In Azure Container Service we've been offering a choice of solutions to our customers for some time. What has this taught us about orchestration technology? What does the future look like? Who is the clear winner? [Spoiler alert: no clear winner but there are clear success stories for each offering.]


Speakers
RG

Ross Gardler

Principal Program Manager, Microsoft


Friday September 15, 2017 10:00am - 10:20am
Diamond Salon 4 & 5

10:20am

11:00am

Practical Container Scheduling: Juggling Optimizations, Guarantees, and Trade-Offs at Netflix - Sharma Podila, Netflix
Assigning resources from distributed clusters to containers from disparate use cases can be tricky. It becomes even more so in practice when combining multiple scheduling objectives and constraints such as bin packing, task locality, and capacity guarantees, among others.

Containers are increasingly used to run microservices, batch, and stream processing applications in large scale shared Mesos clusters at Netflix. This talk dives deep into the challenges, design, and trade-offs achieved using open source scheduling library, Fenzo, that takes a holistic approach, along with plugins based extensibility, to provide a nimble a scheduling core for various independently evolving clusters. Using results, we highlight aspects of capacity guarantees, task placement, elasticity, and operational insights to tackle large scale operations.

Speakers
avatar for Sharma Podila

Sharma Podila

Senior Software Engineer, Netflix
Sharma Podila is a Senior Software Enginner in the Edge Engineering team at Netflix, Inc. His current work includes Fenzo, an open source generic scheduler with plugin based optimizations, developing cloud native Mesos frameworks, and evolving the microservices platform at Netfli... Read More →



Friday September 15, 2017 11:00am - 11:40am
Diamond Salon 6

11:00am

Multi-Tenancy in Apache Mesos - Ben Mahler, Mesosphere & Jay Guo, IBM

This talk will provide an overview of multi-tenancy in Apache Mesos, including recent work that was done to enable multi-tenant frameworks (with multi-role framework support and support for hierarchical roles). Topics include multi-tenant resource management, isolation, security, and other multi-tenancy concerns. This talk will also cover some upcoming work.


Speakers
JG

Jay Guo

Software Engineer, IBM China
I'm working at IBM China as software engineer. I contribute Apache Mesos and HyperLedger Fabric, and I have given three talks on MesosCon during last two years.


Friday September 15, 2017 11:00am - 11:40am
Diamond Salon 7

11:00am

How Yelp.com Runs on Apache Mesos in AWS Spot Fleet for Fun and Profit - Kyle Anderson, Yelp
AWS Spot Fleet is a mechanism to launch EC2 instances in AWS with deep discounts (~80%).
They come with a catch: instances can be recalled with a 120 second notice.

Can the business value of saving all that money be balance with the operational risk of running on such (theoretically) unstable infrastructure? At Yelp the answer to that question is "Yes!"

In this presentation Kyle Anderson from Yelp will explain the how Yelp.com runs on this infrastructure, including the special configuration tweaks needed for Mesos, Marathon, Chronos, maintenance primitives, and custom autoscalers, that are necessary to run keep the infrastructure reliable. (Spoiler alert: Mesos is the key)

Speakers
KA

Kyle Anderson

Site Reliability Engineer, Yelp
Kyle Anderson is a Site Reliability Engineer working at Yelp. | | He helps build and run "PaaSTA", Yelp's open source platform-as-a-service built on Mesos, running on a hybrid infrastructure composed of AWS and bare metal servers.


Friday September 15, 2017 11:00am - 11:40am
Diamond Salon 3

11:00am

Accelerating Spark Workloads in an Apache Mesos Environment with Alluxio - Adit Madan, Alluxio, Inc.
Organizations Mesos and Apache Spark together to gain insight from large amounts of data. It is common for Spark to process data stored in disparate public cloud storage, such as Amazon S3, Microsoft Azure Blob Storage, or Google Cloud Storage as well as on-premise data on HDFS, Ceph or ECS. This architecture results in sub-optimal performance as data and compute are not co-located.

Using Alluxio, an open-source memory speed virtual distributed storage system, deployed on Mesos enables connecting any compute framework, such as Apache Spark, to storage systems via a unified namespace. Alluxio enables applications to interact with any data at memory speed. Alluxio can eliminate the pains of ETL and data duplication, and enable new workloads across all data. Adit will discuss the architecture of Mesos, Spark and Alluxio to achieve an optimal architecture for enterprises.

Speakers


Friday September 15, 2017 11:00am - 11:40am
Diamond Salon 1 & 2

11:50am

Simulating Our Apache Mesos Framework for Fun and Profit and... - Wil Yegelwel, TwoSigma
At Two Sigma, we use Cook, our open source batch scheduling mesos framework, to run millions of compute-hours of work for hundreds of users every day. This year, we are working on changing the scheduler optimization algorithm, but we want some confidence that it will actually improve utilization and the user experience (latency). Therefore, we built a mesos simulator which lets us test algorithm changes, without running the whole distributed system, and in doing so, found some subtle bugs and performance bottlenecks, and highlighted areas needing better test coverage.

In this talk, Wil will describe how we built the simulator for cook, some of the challenges in building it, and share the insights we learned by running our historical job traces through it. We will highlight the value and practicality of simulation testing for real, production systems.

Speakers
WY

Wil Yegelwel

Software Developer, TwoSigma
I'm a developer at Two Sigma Investments working on Distributed System and Analysis tools. Previously, I was a student at Brown University studying Computer Science and Applied Math.


Friday September 15, 2017 11:50am - 12:30pm
Diamond Salon 6

11:50am

Improving Storage Support in Apache Mesos using CSI - Steve Wong, {code}; Jie Yu & James DeFelice, Mesosphere; & Chakravarthy Nelluri, Diamanti
Mesos currently supports both local and external persistent volumes. However, both of them have some limitations. For instance, local persistent volume support requires tedious setup by operators for additional disks. External persistent volume support does not support dynamic provisioning, and Mesos doesn’t track the resources associated with the external volumes. This makes quota control, reservation, fair sharing almost impossible to implement. Our goal is to improve the experience for both of them. The solution is based on a general resource provider model in Mesos, and a standard pluggable interface when talking to storage platforms. This interface is called Container Storage Interface (CSI). This talk will give you an overview about CSI and the resource provider based solution for storage, and dive deeper into how storage platforms can be easily integrated with Mesos using CSI.

Speakers
avatar for James DeFelice

James DeFelice

Distributed Applications Engineer, Mesosphere
James is a Tech Lead at Mesosphere, Inc, currently focused on framework development and storage. Before joining Mesosphere, he spent time building on-demand VM provisioning platforms and supporting Mesos users in the wild. James has an MS in Bioinformatics from the Rochester Inst... Read More →
avatar for Chakravarthy Nelluri

Chakravarthy Nelluri

Chief Cloud Architect, Diamanti, Inc.
Chief Architect and founding engineer at Diamanti, the company behind industry's first bare metal container platform. Specializes in Containerization, Networking & Storage.
avatar for Steve Wong

Steve Wong

Strategic Open Source Partner Engineer, {code}
Steve Wong is an Open Source Engineer with the {code} team. Steve has been participating in the Apache Mesos, DC/OS, Kubernetes, and REX-Ray projects.
avatar for Jie Yu

Jie Yu

Tech Lead, Mesosphere
Jie Yu is a Tech Lead at Mesosphere, Inc, focused on containerization, storage and networking. Before joining Mesosphere, he was a software engineer at Twitter. Jie obtained his PhD in Computer Science and Engineering from the University of Michigan where he conducted research fo... Read More →


Friday September 15, 2017 11:50am - 12:30pm
Diamond Salon 7

11:50am

Monitoring Mesos with InfluxDB & the TICK Stack - Paul Dix, InfluxData & Tehmasp Chaudhri, Oracle Data Cloud
In this presentation, Paul Dix will introduce the open source TICK stack, a complete solution for monitoring infrastructure, applications, network, and services. Based on InfluxDB, the open source time series database, the other components of the stack provide everything a developer or operator needs to collect, store, monitor, and visualize monitoring and telemetry data from DC/OS and services and applications deployed within it. Paul will go through an end-to-end example showing how to get full visibility in your infrastructure using nothing but open source tools and software. 

Speakers
avatar for Tehmasp Chaudhri

Tehmasp Chaudhri

Cloud Architect, Oracle Data Cloud
Cloud Architect at Oracle Data Cloud
PD

Paul Dix

Paul is CTO and founder of InfluxData, the company behind the open source time series database InfluxDB. He has helped build software for startups, large companies and organizations like Microsoft, Google, McAfee, Thomson Reuters, and Air Force Space Command. He is the series editor for Addison... Read More →


Friday September 15, 2017 11:50am - 12:30pm
Diamond Salon 3

11:50am

Real-Time Event Processing and Handling Stateful Applications on Apache Mesos - Balajee Nagarajan & Venkatesh Sivasubramanian, GE Digital
In order to deliver DevOps a cloud native application development and deployment experience, most enterprises are seeking to run all workloads on a container runtime platform. Come see us cover some of our experience in running non-trivial workloads on Mesos.
Key Points:
1. Quick intro to industrial use-cases and the problem space
2. Learn about our PaaS Journey (w/ Cloud and CloudFoundry)
3. Discussion on how GE is handling stateful applications on Mesos
4. Big Data with Mesos: Learn about how we built a stream processing platform on Mesos:
. Spark Streaming on Mesos (With near realtime performance)
. Heron/Storm on Mesos (where latency in less than few ms matters)
5. Learn how we managed the complexity of stateful applications on a container platform
. Share our experiences with external persistence
6. Demo
a. See how we stitch all these together for a real world use case.

Speakers
BN

Balajee Nagarajan

Director Software Engineering, GE Digital
Balajee Nagarajan is currently Director of Software Engineering at GE Digital, working on building the Predix Platform, where he architects and builds the Container and Application runtime platforms. Prior to joining GE Digital, Balajee was part of the Avi Networks team building... Read More →
VS

Venkatesh Sivasubramanian

Venkatesh Sivasubramanian is currently a Senior Director at GE Digital, where he drives the architecture and development of Data Services for Predix, an Industrial IoT platform. Prior to joining GE Digital, he worked as a lead engineer in the Big Fast Data team at WalmartLabs, bu... Read More →



Friday September 15, 2017 11:50am - 12:30pm
Diamond Salon 1 & 2

12:30pm

Lunch (Attendees on Own)
Friday September 15, 2017 12:30pm - 2:00pm
TBA

2:00pm

Automated Performance Tuning with Bayesian Optimization - Joshua Cohen & Ramki Ramakrishna, Twitter
Managing resource utilization is one of the hardest aspects of operating Twitter’s Mesos clusters. As the number of services grows and their resource shapes diversify, the bin packing problem becomes increasingly difficult. Tuning for optimal performance would reduce resource usage, and ease the bin packing burden. However, the multitude of available knobs, heterogeneous hardware, the large number of services, and software and hardware upgrades together make the tuning problem combinatorially intractable.

At Twitter we are developing a system that continuously performs automated tuning of services running in our Mesos clusters, using a machine learning technique called Bayesian optimization. This technique allows us to efficiently search very large parameter spaces to optimize specific performance metrics. We describe our system and share initial results.

Speakers
JC

Joshua Cohen

Twitter, Twitter
Joshua Cohen is a Senior Software Engineer at Twitter on the VM Team, working on performance optimization of JVM services. He is also a committer and PMC member for the Apache Aurora project where he has focused on deploy tooling and filesystem isolation. Previously, amongst othe... Read More →
RR

Ramki Ramakrishna

Staff Software Engineer, Twitter
Ramki Ramakrishna is a staff software engineer in the Infrastructure Engineering Division of Twitter. He is a member of the JVM Platform team and of the Twitter Architecture Group. Ramki has worked with several generations of the JVM, at Sun and Oracle, before Twitter. He has bee... Read More →


Friday September 15, 2017 2:00pm - 2:40pm
Diamond Salon 6

2:00pm

Expanding Mesos Ecosystems: Enabling Mesos on Windows - Andy Schwartzmeyer & Jeff Coffler, Microsoft
Microsoft has worked closely with Mesos and DC/OS communities to support customer workloads on Windows by expanding Mesos Ecosystem. In this presentation, we will review some of our work and learnings, then focus on the enhancement on CMake build system providing a single cross-platform developer experience, also discuss the roadmap for what’s coming next and show some cool demos!

Speakers
avatar for Jeff Coffler

Jeff Coffler

Sr. Software Engineer, Microsoft
Jeff Coffler is a Sr. Software Engineer at Microsoft based in the Seattle area. He has over 20 years of experience working on high performance, cross-platform UNIX/Linux software. Currently, he’s working on full support of Mesos on the Windows platform. In his spare time, he en... Read More →
avatar for Andrew Schwartzmeyer

Andrew Schwartzmeyer

Software Engineer, Microsoft Corporation
Andrew Schwartzmeyer is a cross-platform software engineer and open-source evangelist at Microsoft, where he leads the effort to bring full Windows support to Mesos. Andrew has been an open-source contributor since his college days, and previously brought PowerShell to Linux. In... Read More →



Friday September 15, 2017 2:00pm - 2:40pm
Diamond Salon 7

2:00pm

Astronomer's Journey from Cloud Lock-in to OSS Independence - Aaron Brongersma, Astronomer
Astronomer is a data engineering platform that collects, processes and unifies users' data so that they can get straight to analytics and data science. Initially, Astronomer built their pipeline out of AWS-managed services. But, when they ran into the limitations of these services, they realized that they had committed to proprietary components, which prevented them from changing cloud providers.

In this talk, Astronomer outlines their transition to open source services running on Apache Mesos and DC/OS. They started by using Apache Airflow instead of AWS Simple Workflow Service and went on to replace their Kinesis streams with Apache Kafka, and deploy their stream processing applications with Marathon. They will review their remaining challenges and the solutions they are considering. It took some time and effort to for Astronomer decouple from AWS, but they are glad they did.

Speakers
avatar for Aaron Brongersma

Aaron Brongersma

VP of Engineering, Astronomer



Friday September 15, 2017 2:00pm - 2:40pm
Diamond Salon 3

2:00pm

Distributed Deep Learning on Apache Mesos with GPUs and Gang Scheduling - Min Cai, Alex Sergeev, Paul Mikesell & Anne Holler, UBER
Distributed deep learning is essential to speed up complex model training, scale out to hundreds of GPUs, and shard models that can not be fit into a single machine. With recent advance on deep learning models in self-driving car areas such as lane-detection, perception and so on, it is important to enable distributed deep learning with large-scale GPU clusters.

This presentation will discuss our design and implementation of running distributed TensorFlow on top of Mesos clusters with hundreds of GPUs. It leverages several key features offered by Mesos such as GPU isolation and nested containers. We also implement several features in our scheduler to support GPU and Gang scheduling, task discovery and dynamic port allocation. Finally, we will show the speed up of distributed training on Mesos using an example TensorFlow model for image classification.

Speakers
MC

Min Cai

Staff Engineer, UBER
Min Cai is a Staff Engineer at UBER working on cluster management. He received his Ph.D. degree in Computer Science from USC. Before joining Uber, he was a Sr. Staff Engineer at VMware working on vMotion and vSphere.
AS

Alex Sergeev

Senior Engineer, UBER
Alex Sergeev is a Senior Engineer at UBER working on scalable Deep Learning. He recived his MS. degree in Computer Science from MEPhI. Before joining UBER, he was Senior Engineer at Microsoft working on Big Data Mining.


Friday September 15, 2017 2:00pm - 2:40pm
Diamond Salon 1 & 2

2:50pm

8 Tips for Marathon Performance - Tomasz Janiszewski, Allegro
Mesosphere Marathon is a Mesos scheduler that can handle huge production installation. In this presentation, I will present 8 tips that improves Marathon performance and prevents outages. Each tip will have summary with information when and how it can decrease performance and how to avoid the danger of the outage.

Speakers
avatar for Tomasz Janiszewski

Tomasz Janiszewski

Software Engineer, Allegro
Tomasz is a software engineer passionate about distributed systems. He believes in free and open source philosophy and occasionally contributes to projects on GitHub. At Allegro he works as a Software Engineer working with Mesos and Marathon cluster.



Friday September 15, 2017 2:50pm - 3:30pm
Diamond Salon 6

2:50pm

Challenges of File System Isolation - Santhosh Kumar Shanmugham, Twitter
Twitter use Apache Aurora on Apache Mesos to provide a scalable cluster that is used by the engineers to run their microservices. The platform has seen dramatic adoption due to the deploy features provided by Aurora, which is appealing to the engineers and the resource isolation features provided by Mesos, which improves the cluster utilization. At Twitter we use isolation of CPU, Memory, Disk and Network bandwidth to provide guaranteed access to resources for application running inside a container. One of the glaring exceptions of the container isolation is the lack of file system isolation. To this end we experimented with enabling Docker containers inside Twitter to provide an isolation layer between the host and the container. In this talk we will share our experiences at an attempt to using Docker at scale and the lessons we have learned along the journey.

Speakers
avatar for Santhosh Kumar Shanmugham

Santhosh Kumar Shanmugham

Sr. Software Engineer, Twitter
I am a Software Engineer at Twitter working on Apache Aurora and Apache Mesos projects that provides Twitter's Compute Platform. Recently I have been working on enabling filesystem isolation via Docker containers at Twitter.



Friday September 15, 2017 2:50pm - 3:30pm
Diamond Salon 7

2:50pm

Using External Persistent Volumes to Reduce Recovery Times and Achieve High Availability on DC/OS - Dinesh Israni, Portworx Inc.

Most modern distributed applications like Cassandra and HDFS provide replication of data across nodes and failure zones to be able to deal with failures. But the time taken to recover to a pre-failure level of redundancy in cases of permanent node failures can be large, since a lot of data needs to be copied over to the new node. Also, some of these applications cannot accept new writes on the nodes being bootstrapped, further increasing the recovery time.


Speakers
avatar for Dinesh Israni

Dinesh Israni

Senior Software Engineer, Portworx Inc
Dinesh Israni is a Senior Software Engineer at Portworx with over 7 years of experience building Distributed Storage solutions. Prior to Portworx, Dinesh was at Microsoft, through their acquisition of StorSimple, working on their Hybrid Cloud Storage solution. Recently, he has be... Read More →



Friday September 15, 2017 2:50pm - 3:30pm
Diamond Salon 3

2:50pm

Distributed Data on Distributed Infrastructure - Jörg Schad, Mesosphere; Claudius Weinberger & Kunal Kusoorkar, ArangoDB
Speakers
avatar for Jörg Schad

Jörg Schad

Software Engineer, Mesosphere
Jörg is a software engineer at Mesosphere in Hamburg. In his previous life he implemented distributed and in memory databases and conducted research in the Hadoop and Cloud area. His speaking experience includes various Meetups, international conferences, and lecture halls.
avatar for Claudius Weinberger

Claudius Weinberger

ArangoDB GmbH, CEO



Friday September 15, 2017 2:50pm - 3:30pm
Diamond Salon 1 & 2

3:30pm

Afternoon Coffee Break
Friday September 15, 2017 3:30pm - 4:00pm
Diamond Level - Foyer

4:00pm

OpenWhisk as an Apache Mesos Framework - Tyson Norris & Dragos Dascalita Haut, Adobe Systems, Inc.
OpenWhisk is quickly gaining momentum as a serverless platform, where function developers can deploy their executable code (or even docker containers), without having to manage builds or servers or monitoring or all the things required to host a conventional web application. To date, it has not been possible to leverage the cluster-wide scheduling facilities for running functions via OpenWhisk. We will present a Mesos framework for efficiently leveraging cluster resources for use by OpenWhisk functions, and trace the evolution of this framework from web application, to Mesos framework, and finally to DC/OS package based on dcos-commons.

Speakers
avatar for Dragos Dascalita Haut

Dragos Dascalita Haut

Project Lead, Adobe I/O, Adobe Systems, Inc.
Dragos Dascalita Haut is working as Project Lead for Adobe I/O, working with his team in Romania, India and his hometown San Jose on scaling Adobe’s APIs using a distributed serverless platform. Dragos prefers dynamic programming languages over static ones, dogs over cats, and... Read More →
TN

Tyson Norris

Adobe Systems, Inc.
Tyson Norris is a Sr. Computer Scientist at Adobe Systems, where he works on the Adobe I/O Runtime project. He has been building web applications for 20 years, at Adobe for the last 7.


Friday September 15, 2017 4:00pm - 4:40pm
Diamond Salon 6

4:00pm

Idle Resources Oversubscription: Improving Cluster Utilization - Dmitry Zhuk, Twitter
In this talk we will focus on effective Mesos cluster utilization for non-production jobs, which are commonly used for development and testing. While many of such jobs are idle, they consume cluster resources, leaving a huge gap between allocated and actually utilized resources.
We will present our approach for detecting idle resources, offering them to other non-production jobs using Mesos oversubscription model, and isolating non-production jobs from production jobs. We will also highlight issues in Mesos’ approach to oversubscription and discuss possible solutions and workarounds.

Speakers
DZ

Dmitry Zhuk

Software Engineer, Twitter
Dmitry Zhuk is Software Engineer at Twitter, working in Mesos/Aurora Team.


Friday September 15, 2017 4:00pm - 4:40pm
Diamond Salon 7

4:00pm

“One Ring to Bind Them” - Packet, Qualcomm, & Mesos: A Window into the Development of the Armv8 Ecosystem - Ed Vielmetti, Packet

New,​ ​innovative hardware​ ​brings​ ​with​ ​it​ ​the​ opportunity for architectural disruption as well as the ​challenge of how to effectively tap into a vast existing ecosystem of software.  

Cloud native technologies like DC/OS have only increased the velocity and importance of software.  And yet the value to be unlocked from new architectures is ever more compelling, and well worth the struggle.  

Previous MesosCon talks (Asia, June 2017) have spoken about running Mesos on Armv8 processors.  Packet’s Ed Vielmetti will ​use the Qualcomm Centriq™ 2400 (the world’s first 10nm server processor), as a window into the benefits and challenges of embracing new hardware.   

With 48 highly optimized cores, the system provides a compelling new resource for software like DC/OS that is trying to help applications and users get the most out of their infrastructure. Furthermore, Qualcomm’s efforts in concert with the community provides an excellent window into the ecosystem development journey.

As J.R.R. Tolkien said: “All's well that ends better.”



Speakers
avatar for Ed Vielmetti

Ed Vielmetti

Special Projects Director, Packet
Ed is an Internet veteran with over 30 years experience.He has extensive experience with networks at all levels - physical, logical, technical, social, political, and financial. He is a graduate of the University of Michigan, and an alumnus of Cisco Systems and Arbor Networks. A... Read More →



Friday September 15, 2017 4:00pm - 4:40pm
Diamond Salon 3

4:00pm

Running a Cloud-Native SQL Database on Mesos - Ben Darnell, Cockroach Labs
Compute management solutions like Mesos have rapidly advanced in recent years, but storage options have been left behind. Conventional SQL databases still require mountains of development and operational overhead to keep them functional as scale and availability requirements grow. What could be possible if a strongly-consistent database built specifically for cloud-native deployments paired with a powerful orchestration system like Mesos?

This talk will look at how to deploy and manage CockroachDB, a scalable and strongly-consistent database, using sophisticated orchestration tools like Mesos. We will explore the ease with which CockroachDB was configured to run on Mesos using the new DC/OS SDK for stateful services, which saved days of writing code. We will then explore how you can deploy CockroachDB on Mesos to ship scalable, resilient applications on any cloud infrastructure.

Speakers
avatar for Ben Darnell

Ben Darnell

CTO, Cockroach Labs



Friday September 15, 2017 4:00pm - 4:40pm
Diamond Salon 1 & 2

4:50pm

Fault Tolerant Frameworks - Making Use of CNI Without Docker - Aaron Wood & Tim Hansen, Verizon
While most people use Docker for their containerization efforts, it is in no way a requirement for Mesos frameworks. This presentation will cover how Verizon built their own highly available, performant framework utilizing the universal containerizer and the container network interface. Specific topics of interest include how high availability was built, the reasons behind Go as the language of choice, the benefits that CNI provides, pros and cons of using the universal containerizer, and a quick overview of the accompanying SDK. Usage of the new V1 streaming API will also be covered.

Speakers
TH

Timothy Hansen

Tim has been writing software professionally for 4 years, and currently works with the research and development team at Verizon Labs. Topics of professional interest include machine learning and distributed computing. Languages primarily utilized at work are Go, Java, Python, and... Read More →
avatar for Aaron Wood

Aaron Wood

Principal Software Engineer, Verizon
Aaron, a passionate software engineer with over 10 years experience, works on distributed services and cloud architecture at Verizon Labs. He has an affinity for Linux and security and works with Go, C++, Java, and Python in his current role.


Friday September 15, 2017 4:50pm - 5:30pm
Diamond Salon 6

4:50pm

An Overview of Mesos Containerization and the Default Executor - Gilbert Song & Anand Mazumdar, Mesosphere

Mesos, different from some other container orchestrators, has its own native container runtime. Based on the pure Linux Kernel namespaces and cgroups, Mesos supports different container image formats with advantages of extensible container storage, networking, and security.

As of today it support all major container image formats such as OCI or Docker Image format, as well as default storage and networking plugins formats with CSI and CNI.

Historically, Mesos has provided a command executor for running one off tasks by schedulers e.g., Marathon. For launching a group of co-located tasks, the recommended way was to implement the logic using a custom executor. With the support for task groups aka Pods in Mesos, the default executor is the new recommended way for running pods in production. Moreover, based on nested containers, debug containers become realistic to enter any containers namespaces.

In this talk you will learn the following:

- Overview of the Mesos Agent API’s for nested containers used by the default executor

- Best practices for running sidecar/adapter containers and transient tasks

- Health Checks and Probes (Non-interpreted health checks)

- Default Termination Policy for the default executor

- New planned upcoming features on the roadmap


Speakers
AM

Anand Mazumdar

Distributed Systems Engineer, Mesosphere
Anand Mazumdar is an Engineer at Mesosphere and a committer on the Apache Mesos project. Prior to that, he used to work at a Quantitative Hedge Fund and Amazon Web Services on scalable data stream processing. He holds a Masters in Computer Science from The University of Texas at... Read More →
avatar for Gilbert Song

Gilbert Song

Distributed Systems Engineer, Mesosphere
Gilbert Song, Apache Mesos PMC/Committer, is a Distributed Systems Engineer at Mesosphere. He has been contributing to Mesos for years and mainly focuses on Mesos Containerization. He holds a Master’s degree in Computer Engineering from University of California, Santa Barbara... Read More →


Friday September 15, 2017 4:50pm - 5:30pm
Diamond Salon 7

4:50pm

Doing “DropBox” the Cloud Native Way - Jorg Schad, Mesosphere
Cloud Native architecture has slowly become the default way to build robust, scalable applications. How would you rebuild a large scale storage service such as Dropbox (please note Dropbox is just an example here and we have no plans of actually competing with Dropbox), using the Cloud Native architecture? In this presentation, Joerg and Nitish will discuss the Cloud Native architecture, its advantages, and then explain how to build a scalable, multi-tenant, Dropbox like cloud storage system using modern, containerized applications like Minio as the unstructured data/blob store, MongoDB for structured data, Redis for key value pairs etc, all orchestrated on DC/OS. The presentation will also include a live demo of the Dropbox equivalent product.

Speakers
avatar for Jörg Schad

Jörg Schad

Software Engineer, Mesosphere
Jörg is a software engineer at Mesosphere in Hamburg. In his previous life he implemented distributed and in memory databases and conducted research in the Hadoop and Cloud area. His speaking experience includes various Meetups, international conferences, and lecture halls.



Friday September 15, 2017 4:50pm - 5:30pm
Diamond Salon 3

4:50pm

Streaming Data Pipelines on Apache Mesos: Lessons Learned - Dean Wampler, Ph.D., Lightbend
Lightbend Fast Data Platform (http://lightbend.com/fast-data-platform) is a suite of open-source streaming tools, including Kafka, Spark, Flink, Akka Streams, and Kafka Streams, running on Mesosphere DC/OS. Lightbend picked Mesos for its flexibility to support the diverse requirements of streaming and batch data processing, as well as other microservices typically required in production environments.

Dean will begin by describing the features of Mesos that were important for Lightbend's needs. He will illustrate the details of those features using examples from how the tools mentioned above work with Mesos, both for long-running daemons and per-job processes. Dean will share lessons learned running applications built with these tools on Mesos and he will conclude with thoughts about future developments and improvements.

Speakers
DW

Dean Wampler

Big Data Architect, Lightbend
Dean Wampler, Ph.D. (@deanwampler), is Vice President of Fast Data Engineering at Lightbend. He leads the development of Lightbend Fast Data Platform, a streaming data and microservices platform built on Lightbend Reactive Platform, Spark, Flink, Kafka, Mesosophere DC/OS, and oth... Read More →


Friday September 15, 2017 4:50pm - 5:30pm
Diamond Salon 1 & 2