AWS Community Days

 

 

Frankfurt | 26 & 27 September 2018

 

 

#AWSCommunity

The registration for this event has closed. Thank you for your interest.

Presentations

The session slides are now available. You'll find the links below at Detailed Agenda.

What are the AWS Community Days?

The AWS Community Days are a 2-day event for AWS users organized by the communities themselves. Once a year all AWS users come together to network, learn from each other and get inspired by deep-tech sessions presented by expert AWS users and community leaders.

Tracks

Agenda

Overview

26 September

12:00 - 13:00 Registration and Lunch
13:00 - 13:45 Community Welcome and Kick-Off
14:00 - 17:45 4 Tracks including 4 x 45 min. Sessions
From 18:00 Networking Evening incl. Dinner Buffet and Drinks
4 Tracks including 4 x 45 min. Sessions

27 September

08:00 - 09:00 Registration & small Breakfast
09:00 - 11:45 4 Tracks including 3 x 45 min. Sessions
11:45 - 13:00 Lunch Break
13:00 - 14:45 4 Tracks including 2 x 45 min. Sessions
4 Tracks including 4 x 45 min. Sessions
4 Tracks including 3 x 45 min. Sessions

Detailed Agenda

  • Serverless Computing

    Track Owner: Thorsten Höger (@hoegertn - UG Stuttgart + UG Frankfurt)

    26 September
    14:00 - 14:45

    View Presentation »

    Production-ready Serverless Java Application in AWS Cloud within 3 weeks
    Vadym Kazulkin & Elmar Warken, ip.labs GmbH

    In unserem Vortrag tauchen wir tiefer in die Serverless-Welt ein und zeigen wie eine produktionsreife Java Serverless-Anwendung mithilfe von AWS-Cloud mit dem Technologie-Stack API Gateway, SNS, Lambda and DynamoDB aufgebaut werden kann. Dabei gehen wir auf die Besonderheiten der jeweiligen Services ein, wie "cold start" bei Lamda oder "provisioned throughput" bei DynamoDB. Dabei zeigen wir, welche Strategien und Wege es gibt, um damit umzugehen. Am Ende werfen wir einen Blick in die Zukunft und sprechen über die erste relationale serverless Datenbank "Aurora Serverless".

       

    26 September
    15:00 - 15:45

     

    Serverless Product Recommendations using AWS Recognition 
    Olalekan Elesin, Scout24

    AWS Rekogniton makes it very easy to build image and video analysis into your applications without prior deep learning skills. As the amount of image and video data continues to explode, engineers should consider leveraging cloud economies of scale to harness the potentials of this explosion. Amazon Rekognition provides developers capacity to wield this power with minimal or no knowledge of deep learning. Software engineers building deep learning capabilities into their applications are often buried in acquiring knowledge about the field, perfecting models and managing the pipeline infrastructure. In this talk, we will draw inspiration from product recommendations (a popular use case for machine learning) and design a serverless architecture that allows us to integrate image recommendations into an application without prior knowledge of deep learning, model fitting and infrastructure management.

       

    26 September
    16:00 - 16:45

    View Presentation »

    Your Lambda function might execute twice. Be prepared!
    Michael Wittig, widdix GmbH

    Are you confused when scheduled Lambdas execute twice, SNS messages trigger an invocation three times, your handmade S3 inventory is out of sync because events occurred twice? Bad news: Sooner or later, your Lambda function will be invoked multiple times. You have to be prepared! The reasons are retries on errors and event sources that guarantee at-least-once delivery (e.g., CloudWatch Events, SNS, …).In this talk, you will learn why Lambda functions might execute twice and what you can do about it.

       

    26 September
    17:00 - 17:45

    View Presentation »

    Serverless Architecture meets Continuous Delivery
    Andreas Mohrhard, cosee

    Serverless Architekturen basieren auf Function-as-a-Service-Angeboten, wie zum Beispiel AWS Lambda, welche den Kleber für Serverless Ressourcen darstellen. Genauer gesagt ist AWS Lambda eine elegante und skalierbare Lösung für die Verarbeitung von Events in der Cloud. Geeignete Beispiele sind die Analyse von eingehenden E-Mails und das Konvertieren von Dateiuploads in andere Formate. In Kombination mit serverless API Gateways eignet sich AWS Lambda auch für den Betrieb von HTTP-basierten APIs.

    Aus Entwicklerperspektive sollten Serverless Architekturen einfacher zu ändern sein, da das Pflegen der zugrundeliegenden Infrastruktur entfällt. Hierbei wird häufig übersehen, dass Serverless Applikationen meist mehrere Cloud-Dienste und zahlreiche Cloud-Ressourcen umfassen und daher eine manuelle Konfiguration dieser stark fehleranfällig ist. Dem neuesten Stand der Technik entsprechend ist die Lösung für dieses Problem Continuous Integration/Delivery (CI/CD). Jedoch lassen sich etablierte Praktiken und Tools für CD-Pipelines nicht auf natürliche Art auf Serverless Architekturen anwenden und schaffen hierdurch neue Herausforderungen.
    Wir zeigen Lösungen für die bei der Entwicklung von Serverless Anwendungen zwangsläufig auftretenden Probleme auf. Hierbei diskutieren wir Methoden, etablierte Praktiken und Tools für die Realisierung von Serverless CI/CD-Pipelines für AWS Lambda und verwandte Dienste. Demonstriert werden diese anhand einer Single Page-Applikation, welche ausschließlich auf Serverless Technologien basiert.

       

    27 September
    9:00 - 9:45

    View Presentation »

    Assembling production-ready applications using Serverless App Repository
    Aleksandar Simovic, AWS Serverless Hero

    Serverless computing provided us with many great small mechanisms to build and scale our applications. AWS Serverless Application Repository extended that even more, by enabling developers to share their serverless application "recipes" and provide serverless solutions out-of-the-box, amplifying the speed of development to a matter of days, not sprints. However, you need to know how to assemble these solutions together along with your specific business logic and deliver production-ready applications. In this session, we will show you how to design and combine composite serverless applications using AWS Serverless Application Repository.
    Learn about potential problems and caveats you might encounter, how to "combine two pieces of a puzzle that don't fit together" and what can you rely on.

       

    27 September
    10:00 - 10:45

     

    Incorporate testing and security into your serverless DevOps pipeline
    Mahdi Azarboon, Accenture

    Serverless computing is getting more popular and so are its challenges. DevOps and testing are among the most notable issues, and this has made many serverless users to neglect importance of DevOps. It’s common to hear that users deploy their apps manually and perform just few basic unit tests. Serverless community should work better to implement DevOps principles. Otherwise, users will pay repeatedly for this lack-of-automation over life cycle of their projects. Meanwhile, multi-clouding is getting more popular, that’s why users need a vendor agnostic tool to perform continuous operation and testing. How to create a vendor-agnostic and flexible CI/CD pipeline for your serverless app? How to automatically perform unit, integration and security tests into your pipeline? These are the questions that Mahdi will answer in his session.

       

    27 September
    11:00 - 11:45

    View Presentation »

    Data Processing using Serverless
    Nils Rhode, Haufe Lexware

    Haufe Group – with its brands Haufe, Haufe Akademie, and Lexware – is considered a pioneer of digital transformation in Germany and one of Germany's leading providers of digital workplace solutions and services, as well as offering education and training programs. The content engineering and development (CED) division of Haufe is responsible for internal applications and products for the target groups (e.g. mid- and large-size companies and consultants, layers). As part of the ongoing initiative to split the monolith - our Haufe redaction system (HRS) – into microservices, the HRS team at Haufe.Group started with the first component (EDB - court decision database), taking a serverless computing approach. Nils will present the serverless solution for ingest and export of content to and from the EDB using serverless workflows. In this session, he address why the HRS team chose to go serverless with the EDB and dives into the technical solution developed by the HRS team. This includes the serverless architecture, how to write serverless functions and how to orchestrate them effectively. 

       

    27 September
    13:00 - 13:45

    View Presentation »

    Medienverarbeitung mit AWS Batch
    Patrick Wolf, Cosee

    Wenn‘s mal wieder länger ...dauern muss - Medienverarbeitung mit AWS Batch Wenn ich Prozesse habe, die aus vielen, kleinen, am besten unabhängigen Arbeitsschritten bestehen und ich mich um nichts kümmern will, ist AWS Lambda für mich das Mittel der Wahl. Aber es gibt die 5-Minuten-Laufzeitbegrenzung. Was mache ich also, wenn ich Prozesse habe, die regelmäßig länger laufen? Vor diesem Problem stehen wir immer wieder, wenn wir Medien wie Filme oder Hörbücher verarbeiten und absichern müssen. Transkodieren, automatisch Previews erstellen, mit Wasserzeichen absichern, das alles sind Arbeitsschritte, die eng an die Abspielzeit des Mediums geknüpft sind und deswegen gerne über die magische 5-Minuten-Marke geraten. In diesem Talk zeigen wir, wie wir dieses Problem mit AWS Batch gelöst haben, so dass es sich von außen betrachtet kaum von einer echten serverless-Lösung unterscheidet. Batch erlaubt uns die Verarbeitung in Einzelschritten abzuhandeln, einfach Abhängigkeiten zwischen diesen Einzelschritten zu definieren und stellt selbstständig sicher, dass einmal angestoßen die ganze Verarbeitungskette durchläuft. Es bleibt allerdings an uns, den Überblick über die angestoßenen Prozesse zu behalten. Dann aber bietet Batch nicht nur die für dieses Szenario nötige Compute-Power, sondern das auch noch zu einem günstigeren Preis als er in Lambda möglich wäre. 

       

    27 September
    14:00 - 14:45

    View Presentation »

    Smart Cat(egorization) with AWS Lambda
    Benjamin Weigel, Europace AG

    Building a Serverless multi-class document classification service using AWS Lambda.

  • Cloud and Software Architecture

    Track Owner: Markus Ostertag (@osterjour - UG München), Lars Wolff (@larsvegasUG Köln) & René Weseler (@rweseler - UG Münsterland)

    26 September
    14:00 - 14:45

    View Presentation »

     

    Alexa, ask compliance checker for my AWS compliance report
    Steffen Mazanek, T-Systems International GmbH

    In this talk we show how an Alexa skill can be provided that informs the user about the compliance status of an AWS account.

    We cover very briefly the basic components of an Alexa skill. We show the Alexa Skills Kit webportal and CLI. Since this skill certainly should be private, we explain how the quite new AWS Alexa Business service can be used to distribute a skill only to a specific set of users.

    Then we show how the new python SDK for Alexa can be used to implement the backend lambda functions. Compliance-related data then can be grabbed from the APIs of TrustedAdvisor, ConfigRules, Cloudtrail etc. using boto3 and periodically written to a DynamoDB, together with data from other sources such as Security Information and Event Management system (SIEM).

       

    26 September
    15:00 - 15:45

    View Presentation »

    Building a SaaS application that is sold in AWS Marketplace
    Dominique Meiländer, Buildsimple

    In this talk we want to give an overview about the creation of a combined EC2 and Serverless Application, that is ready to be sold in AWS Marketplace as a SaaS Subscription. Challenges of including the AWS metering, purchasing and integration model of AWS Marketplace. Selling an API-Gateway vs. SaaS Subscription.

       

    26 September
    16:00 - 16:45

    View Presentation »

    Programmed Governance and how to deliver secure and compliant accounts with a click
    Nico Herzhauser, T-Systems International GmbH

    AWS Organizations lets you create and manage accounts for yourself, your business units or your customers. But how can you make sure that these accounts become and stay compliant with your organizational requirements? In this session we will demonstrate and share the lessons learned building a 1-click engine, that creates, provisions and audits compliant account structures using only AWS tools such as step functions, cloudformation stacksets and a little bit of lambda.

       

    26 September
    17:00 - 17:45

    View Presentation »

    How to use or not use the AWS API Gateway for Microservices
    Martin Merck, Cloudarchitect, Cloudreach

    In this talk, we examine how you can use AWS API Gateway to implement Microservice Authentication and Authorization with OAuth2 and OpenID connect. This pattern is really useful for serverless REST APIs implemented using AWS native services. On the other hand, you will face challenges implementing classical OAuth2.0 flows with API Gateway for your own microservices. We will discuss in which situations it makes the most sense to use API Gateway to front your microservices and under which circumstances a different API gateway or own security implementation makes more sense.

       

    27 September
    9:00 - 9:45

    View Presentation »

    Distributing Metastore Update to short-lived EMR clusters
    Gavin Perrie, Data Reply

    In a loosely coupled architecture, distributing updates in a reliable manner to resources that may or may not currently be running is a challenge. In this session see how this has been addressed using serverless services to distribute irregularly timed updates to short-lived EMR clusters across multiple timezones.

       

    27 September
    10:00 - 10:45

    View Presentation »

    AWS Landing Zone - How to setup a baseline secure multi-account AWS environment configured based on best practices
    Marc Schröter, globaldatanet

    A landing zone is a secure, scalable, multi-account AWS environment based on AWS best practices. One of those best practices is to have multiple core AWS accounts. One of storing logfiles, one for security purposes, one for automation and one for running shared services like LDAP, DNS or Active Directory. A demonstration of how we provision new accounts and deploy security and compliance features and services to new AWS accounts with a CI/CD pipeline.

       

    27 September
    11:00 - 11:45

    View Presentation »

    Chaos Engineering - withstanding turbulent conditions in production
    Benjamin Wilms, Codecentric

    Die Komplexität in modernen und verteilten Architekturen nimmt immer weiter zu. Wir haben es erfolgreich vollbracht, unsere Applikation in kleine und wartbare Komponenten zu zerlegen. Jede einzelne Komponente lässt sich automatisiert und zu jederzeit in Produktion bringen. Es wurde bei der Entwicklung viel Aufwand betrieben, die Testabdeckung möglichst hoch zu halten. Jedes Release muss erfolgreich durch unsere Pipeline und unzählige Unit-, Integrations- und Akzeptanztests bestehen. Aber warum haben wir kurz vor der Ankunft am schönsten Ort der Welt (Produktion) dieses ungute Gefühl? Viele offene Fragen lassen sich nicht durch einfache Unit- oder Integrationstests beantworten. Hier kommt das Thema Chaos Engineering ins Spiel. Es hilft uns, des Chaos‘ Herr zu werden und bitte behaupten Sie nun nicht, bei Ihnen herrsche kein Chaos! Es gibt eine ganze Industrie, die uns Ticketsysteme verkauft, mit denen wir das Chaos dokumentieren können. Wie man Chaos Engineering einführt, erfahren sie in diesem Talk. Anhand von praxisnahen Beispielen lernen Sie, was alle schief gehen kann. Zum Ende des Talks führen wir in einer verteilten Anwendung ein Chaos Experiment durch. Mit Hilfe des Chaos Monkey for Spring Boot werden wir versuchen, die Anwendung zum Absturz zu bringen. Was dank der implementierten Resilience-Pattern aber nicht gelingen wird.

       

    27 September
    13:00 - 13:45

    View Presentation »

    The Lambda-DoppelHerz: SNS Mobile Pushes at Scale
    Florian Barth, Stocard

    Keeping your mobile users up to date and engaged is one of the key challenges of mobile applications. This talk will give you a peek at how Stocard solved the challenge of delivering millions of mobile pushes to their user base. Florian Barth will take you on a guided tour deep into the core of Stocard's push pipeline - the mythical "Lambda DoppelHerz", a contraption created from AWS Lambda, SNS, and SQS.

       
    27 September
    14:00 - 14:45

    Everybody needs a SecDevOps toolkit
    Jördis Schröder & Sascha Ziesig

    Real-world account security and SecDevOps is not core business for most of us. But as automation is the key for scale we need to make sure to not only understand the implications but also know several techniques in the area of SecDevOps and how to secure our infrastructure in the cloud. This talk will cover ideas of security monitoring & alerting, show architectural patterns focused on security and how to build deployment pipelines with built-in security tests.

  • Internet of Things

    Track Owner: Sven Seiler (@Seiler_IT - UG Dortmund) & Tom Lorenz (@bombadil_tom - UG Dortmund)

    26 September
    14:00 - 14:45

    View Presentation »

    DevOps for IoT at gridX
    Joel Hermanns, gridX

    This talk will illustrate how gridX built a system from the ground up, transfering
    best practices in DevOps from Cloud Computing to the world of IoT in order to deliver features
    more quickly and more reliable to hundreds of devices in the field.

    Joel Hermanns will discuss how certain Cloud Computing and DevOps paradigms can be
    applied to IoT, how this has helped gridX and how AWS services are used as a solid foundation for the whole infrastructure.

       

    26 September
    15:00 - 15:45

    View Presentation »

     

    IoT-based Real World Tracking with AWS
    Omar Tello, sensalytics

    Learn how sensalytics uses AWS Services to run an IoT-based Real World Tracking Platform. Sensalytics uses more than ten different services from the AWS portfolio to manage sensors, receive raw data, merge data, and provide the core product: Their Business Intelligence platform. The challenge of providing real-time information coupled with event-driven notifications across multiple channels was overcome by selecting the AWS platform. With a rapidly growing number of new sensors, AWS Services meets the scaling requirements.

       

    26 September
    16:00 - 16:45

    View Presentation »

     

    How to implement an IoT solution for Grohe
    Andrea Mercanti, Reply

    As a world leader in sanitary fittings, Grohe has launched a family of connected devices: Grohe Sense - spearhead of its digital transformation.
    Leveraging on an AWS cloud native IoT platform built by Reply, Grohe was able to offer a complete water security system for the residential segment to help minimizing the biggest problem in the house: water damage.
    Based on this use case, we will give you an introduction into a State-of-the-Art architecture for an IoT Platform on AWS.

       

    26 September
    17:00 - 17:45

    View Presentation »

     

    AWS IoT Analytics - A hands-on session
    Philipp Sacha, AWS

    AWS IoT Analytics is a fully managed service that makes it easy to perform sophisticated analytics on IoT bulk data. This session will explain the concepts of AWS IoT Analytics. In the following demonstration, an example will be presented how data can be evaluated with AWS IoT Analytics.

  • Machine Learning

    Track Owner: Steffen Mazanek (@smazanek - UG Dresden)

    27 September
    9:00 - 9:45

    View Presentation »

    Extending an Enterprise Content Management System with conversational functions using application-level ML services from AWS

    Markus Oponczewski, fme AG

    We explain how AWS ML Services Comprehend and Translate are used to extend an Enterprise Content Management System with intelligent functions and how Amazon Alexa and Polly improve the usability of such a system.

       

    27 September
    10:00 - 10:45

    View Presentation »

    Deep Learning for Natural Language Processing:
    Could NLP Prevent a Corporate Scandal? And how to bring your own algorithm to AWS Sagemaker

    Franco Arda

    Discover why Computer Vision Revolutionized NLP. NLP and a Very Popular German Professor in Silicon Valley. Case Study and Deep Dive: Could NLP Prevent a Corporate Scandal? How to use AWS Sagemaker to bring your own algorithm to production

       

    27 September
    11:00 - 11:45

    View Presentation »

    Pimp up real time services in your own datacenter: Our story of integrating Deep Learning to existing SaaS products using highly automated processes and services in AWS
    Philipp Sorg, Econda

    At econda we build SaaS products for customers mainly in the e-commerce sector. More and more web shops are looking for better ways to personalize the end user experience using individual content and product recommendations - ideally across all channels. However established algorithms like collaborative filtering are pushed to their limits. It is not possible to further increase the relevance of predictions, and many scenarios are not covered by these algorithms.

    Deep Learning models have the potential to overcome these drawbacks. They are highly flexible in their structure, support real time inference, and are able to ingest any data. We will present details about the Recurrent Deep Learning Model we built to improve the relevance of personalization. It is based on past and current user interactions and is integrated in web shops, apps or newsletters based on real time services. We will also present success stories of how we increased revenue of our customers based on this feature.

    Developing the Deep Learning model is only one part of the story - integrating new technology in existing infrastructure is another challenge. We will present our solution that includes hops between data centers even for real time services. At AWS, we use Sagemaker for a daily, fully automated training on GPU instances and ECS for the real time inference layer. Everything is glued together by Lambda functions which enable fully automated activation, training, deployment and auto scaling for new customers.

       

    27 September
    13:00 - 13:45

    View Presentation »

    State machine based analytics of big data event streams for an IoT use case
    Krishna Kalyan, zoi

    Connected devices or apps create a huge amount of event data. To derive intelligent actions from this never ending stream of information one has to build a robust decision model. State machines, widespread in embedded development or business systems, provide a perfect foundation for that. We show, how to connect such a pipeline to AWS IoT gateway, use Kinesis to stream the data to a containerized analytical component running on AWS Fargate and use S3/dynamo to store the state. 

       
    27 September
    14:00 - 14:45
    Setup and Object Detection with AWS DeepLens
    Julien Simon, AWS
  • Container/Kubernetes/EKS

    Track Owner: Andreas Rütten (LinkedIn ProfilUG Hamburg) & Philipp Garbe (@pgarbe)

    26 September
    14:00 - 14:45

    View Presentation »

    Pursuit of Stability - Growing AWS ECS in production
    Alexander Köhler, inovex

    When running container at scale there are lot of solutions these days: self hosted k8s, EKS, Openshift, Nomad, Docker Swarm and another multitude of tools. One of those is the AWS Elastic Container Service. In this talk we will look at the history of an ECS cluster that we build for a customer project. We will show its beginnings, how we grew it, how we nearly broke it, how we added additional tooling, how we rebuild its foundation, how we did made our deployment multi-account capable.

       

    26 September
    15:00 - 15:45

    View Presentation »

    How to run Kubernetes on AWS
    Tobias Schwab, PhraseApp

    Kubernetes on AWS in 2018. PhraseApp has been running on AWS since 2014 and even though we migrated all of our deployment infrastructure to Kubernetes in 2017, long before there was any kind of managed Kubernetes on AWS, we had good reasons not to leave. 
    Having started with a self-managed "good enough for the time being" Kubernetes deployment we finally decided to use Kops, an open source tool that allows you to manage production-grade high-availability Kubernetes clusters.
    This talk will give you an introduction to Kops, specifically created for operating  Kubernetes clusters on AWS. By building on rock solid components such as AutoScaling Groups, EBS-Volumes and the ElasticLoadBalancer, Kops allows seamless scaling and fully automated zero-downtime upgrades of your Kubernetes clusters.
    With EKS, the Elastic Container Service for Kubernetes coming to more and more regions it also makes sense to compare Kops with EKS to see what both options have to offer and why Kops could still be the better tool for the job.
    You will also learn how Kubernetes can fit into the other parts of the AWS ecosystem and where it makes more sense to use a fully managed AWS services instead of running things in your Kubernetes cluster.

       

    26 September
    16:00 - 16:45

     

    Kubernetes @ MOIA: cluster-sharing for ride-sharing
    Moritz Siuts, MOIA

    We want to show how our new ride-sharing platform looks like in AWS. Our platforms spans both worlds - Kubernetes & Serverless - and we want to dive into our current setup and technology stack. We will present how we deploy our Kubernetes clusters and what components and services make it useful for us.

       

    26 September
    17:00 - 17:45

    View Presentation »

    reBuy two year kubernetes story
    Stephan Lindauer, rebuy

    The cloud transformation for reBuy is based on two pillars, AWS and
    kubernetes.
    Since we started to move to the cloud two years ago we built everything on top of them.This talk will cover the following topics:
    - The setup we are currently running
    - Things we stopped doing after trying
    - Lessons learned running kubernetes on AWS
    - Our recommendations for going the kubernetes route

       

    27 September
    9:00 - 9:45

    View Presentation »

    Run Jenkins as Managed Product on ECS
    Philipp Garbe, Scout24

    You build it, you run it. But how do you deploy it? Does one CI system fits all of your teams? Why is continuous delivery so important? What's different when you treat the platform as a product?

    In this talk, Philipp shows how Scout24's platform team provides Jenkins as a managed product and the challenges they had. You'll see how Docker can be used to provide immutable build agents and how they leveraged Amazon ECS to provide infinite scaling of agents.

       

    27 September
    10:00 - 10:45

    View Presentation »

    Using Spot Instances with Kubernetes
    Rodrigo Reis, Zalando

    With the purpose of being more cost efficient we at Zalando started evaluating using Spot Instances with Kubernetes. In this session I am going to talk about what we've done, the benefits achieved, but also the problems we currently have and how we solved some.

       

    27 September
    11:00 - 11:45

     

    Cloud-native Microservices
    Timo Derstappen, Giantswarm

    Microservice Architectures are around since a while now. There have already been a few generations of tooling and patterns that helped to run and manage them. Timo will show us how a current cloud-native stack looks like and which tools should be on your radar. The talk will cover both operational and architectural topics. As the name cloud-native suggests most of the projects explained in the talk are either already hosted within the cloud-native computing foundation (CNCF) or within the incubation process. Kubernetes, Prometheus, Envoy, Opentracing and GRPC are just a few examples.

       

    27 September
    13:00 - 13:45

    View Presentation »

    UI-Test Ausführung mit AWS Fargate beschleunigen
    Marcus Pritsch, cosee

    Mit über 400 UI-Tests haben wir eine durschnittliche Testlaufzeit von zwei Stunden. Diese laufen auf einer einzelnen Maschine morgens und abends. Für kritische Hotfixes dauert das Testset aktuell zu lang.
    Alternativen wie Browserstack, um Tests zu parallelisieren und damit die Testlaufzeit drastisch zu verkürzen, sind vergleichsweise teuer.
    Aus unserer Not haben wir uns einer Lösung mit ECS zugewandt, mit der wir parallel Test-Cluster starten können. Diese Cluster lassen sich bei Bedarf skalieren - wie viele Browser hochgefahren werden. Wir präsentieren, wie wir die Orchestrierung der Services (gemocktes Backend, Frontend, Testrunner, Browser) gelöst haben. Außerdem zeigen wir die Einrichtung der Infrastruktur mittels Cloudformation um die Cluster automatisiert deployen und terminieren zu können.

       

    27 September
    14:00 - 14:45

    View Presentation »

    Spinnaker for multi-cloud continuous delivery
    Jukka Forsgren, Nordcloud

    With hyperscale cloud platforms getting more and more attention, the adoption of multi-cloud in big enterprises is starting to be the new norm. While having a sound and scalable platform for your applications is important, the role of efficient delivery pipeline can't be emphasized enough to increase developer efficiency and application development velocity in order to gain competitive advantage. While established tools such as Jenkins and TravisCI can be used quite efficiently for many CI/CD tasks, modern cloud-native tools like Spinnaker are getting more attention and proving to be of high value for large scale continuous delivery. We will show you how Spinnaker can be efficiently utilized in multi-cloud environment together with Kubernetes to speed up your deployments and streamline your application development.

Up-Arrow

Date & Venue

Date: 26 & 27 September 2018

Time: 26.09.18 12:00 to 27.09.18 15:00

Venue: CineStar Metropolis, Eschenheimer Anlage 40, 60318 Frankfurt

Up-Arrow