Not Just Another Mesos Framework

Slides:



Advertisements
Similar presentations
Implementing Tableau Server in an Enterprise Environment
Advertisements

SERVICE BROKER. SQL Server Service Broker SQL Server Service Broker provides the SQL Server Database Engine native support for messaging and queuing applications.
Plan Introduction What is Cloud Computing?
Cloud Computing for the Enterprise November 18th, This work is licensed under a Creative Commons.
Lecture 8 – Platform as a Service. Introduction We have discussed the SPI model of Cloud Computing – IaaS – PaaS – SaaS.
Structured Container Delivery Oscar Renalias Accenture Container Lead (NOTE: PASTE IN PORTRAIT AND SEND BEHIND FOREGROUND GRAPHIC FOR CROP)
Docker for Ops: Operationalize Your Apps in Production Vivek Saraswat Sr. Product Evan Hazlett Sr. Software
Introducing Flink on Mesos Eron Wright – DELL
Open-O SFC.Mgr Proposal
Univa Grid Engine Makes Work Management Automatic and Efficient, Accelerates Deployment of Cloud Services with Power of Microsoft Azure MICROSOFT AZURE.
DevOps Cloud Native Microservices
Containers as a Service with Docker to Extend an Open Platform
Deployment Architectures For Containers
Organizations Are Embracing New Opportunities
4/24/ :07 PM © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN.
Web application hosting with Openshift, and Docker images
Web application hosting with Openshift, and Docker images
Dockerize OpenEdge Srinivasa Rao Nalla.
Introduction to Distributed Platforms
Blue Mixology.
Microsoft Azure-Powered BlueCielo Meridian360 Portal Improves Asset Data Integrity and Facilitates Secure Collaboration with External Stakeholders MICROSOFT.
Docker and Azure Container Service
Running Multiple Schedulers in Kubernetes
Docker Birthday #3.
StratusLab Final Periodic Review
StratusLab Final Periodic Review
Platform as a Service.
CoreDNS and Kubernetes
Distributed Service Bundles
Containers in HPC By Raja.
INDIGO – DataCloud PaaS
Introduction to Cloud Computing
OpenNebula Offers an Enterprise-Ready, Fully Open Management Solution for Private and Public Clouds – Try It Easily with an Azure Marketplace Sandbox MICROSOFT.
9/20/ :55 PM © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN.
Next Gen Microservices
Kubernetes Container Orchestration
Setting policies in kubernetes
Azure Container Service - the most open container orchestration service yet Saurya Das Program Manager.
Azure Container Instances
Oracle DB and Docker Get Your Dockerized Oracle Sandbox Running in the Cloud or On- Premises Martin Knazovicky Dbvisit Software.
Designed for Big Data Visual Analytics, Zoomdata Allows Business Users to Quickly Connect, Stream, and Visualize Data in the Microsoft Azure Platform MICROSOFT.
Securing Cloud-Native Applications Jason Schmitt CEO
HDFS on Kubernetes -- Lessons Learned
Intro to Docker Containers and Orchestration in the Cloud
20409A 7: Installing and Configuring System Center 2012 R2 Virtual Machine Manager Module 7 Installing and Configuring System Center 2012 R2 Virtual.
Kubernetes on Windows Jim Zimmerman Patrick Lang BRK3147
Getting Started with Kubernetes and Rancher 2.0
Intro about Contanier and Docker Technology
Learn. Imagine. Build. .NET Conf
IBM Containers Docker in the Cloud
Abiquo’s Hybrid Cloud Management Solution Helps Enterprises Maximise the Full Potential of the Microsoft Azure Platform MICROSOFT AZURE ISV PROFILE: ABIQUO.
HDFS on Kubernetes -- Lessons Learned
From Source to Production: The Latest in Container Dev
Cloud Web Filtering Platform
Container cluster management solutions
OpenShift vs. Vanilla k8s on OpenStack IaaS
Building and running HPC apps in Windows Azure
Configuration management suite
OpenShift as a cloud for Data Science
Containerized Spark at RBC
OpenStack Summit Berlin – November 14, 2018
SCCM in hybrid world Predrag Jelesijević Microsoft 7/6/ :17 AM
Harrison Howell CSCE 824 Dr. Farkas
A DevOps process for deploying R to production
Deploying machine learning models at scale
Containers and DevOps.
Containers on Azure Peter Lasne Sr. Software Development Engineer
SQL Server on Containers
Presentation transcript:

Not Just Another Mesos Framework Kubernetes on Mesos Not Just Another Mesos Framework Da Ma, Software Architect, kube-mesos owner, IBM Yong Feng, Senior Software Architect, IBM Guangya Liu, Senior Software Engineer, IBM

Why “Kubernetes on Mesos”? Container Runtime The container runtime packages and launches application instance in a sandbox with portable and flexible capability. Docker and rkt are container runtimes. Computing, Storage and Network Application Application Resource Management The resource management component provides an abstraction of resources (cpu, mem, …) for application and then allocates/provision resources among tenants and applications. Mesos is an open source resource manager. EGO is an IBM enterprise resource manager. Workload Management The workload management component manages the life cycle of an application as well as access to the application, including service compose, service discovery, load balance Kubernetes and Marathon are workload managers.

Shared state between fws by Optimistic Offer Why “Kubernetes on Mesos”? Shared state Two Level Scheduling ?? Shared state between fws by Optimistic Offer …… Monolithic PBS/SGE/LSF Mesos/YARN/EGO Res mgr & wld mgr tightly coupled Batch wld only 1992 2003 2016 future

“Kubernetes on Mesos” Organization Sponsor: Tim Hockin (Google) Champion: David Eads (Redhat) Owner: Klaus Ma (IBM) Github: kuberntes-incubator/kube-mesos-framework

Spectrum Conductor with IBM is Owner & User !! Watson/Cognitive Container Cloud Spectrum Conductor with Spark IBM Spectrum Conductor for Container GUI Workflow App Store Authentication Authorization k8s Session Scheduler HPC Persistent Volume Service Discovery Installer (Deploy, Reconfigure, HA, Scale, Rolling update) Mesos Master Distributed Key-value Store Service Load Balance Mesos Agent Trouble-shooting Monitor K8s executor container container pod pod pod container container Network Topology Image registry

Community Value IBM Value-add Customer Value Docker Hub Registry holds a repository of 75000+ Docker images Lots of application integrated with Mesos Kubernetes enable micro-service architecture Client unique registry available on premises Security readiness guidance via the Vulnerability Advisor Build-in applications of popular open source projects and IBM enterprise products in App Store Access to the images and application you require to deploy containers that meet your business needs and strategy Open-source, standardized, lightweight, self sufficient container technology Balance workload between on-prem and off-prem Deployment choice with openPOWER and x86_64 Flexibility to choose on-prem and off- prem or mix for your business Build, ship, and run standardized containers Integrated monitoring & logging Elasticity to grow storage & container needs Integrated CI/CD flow Life-cycle management of containers and data volumes  Docker ease of use combined with enterprise-level integrity and confidence

Promoted to incubator, Sep, 2016 Where are we now? v0.2.0 ~ 0.2.2, 2014 v0.3.0 ~ 0.7.1, 2015 v0.7.2, Jan 30, 2016 Promoted to incubator, Sep, 2016 Re-org the code as a stand along project Removed un-used upstream code Sanity check with latest k8s version Updated CI script for daily build Enabling e2e test

No code dependency Get Pods Match Pods and Offers Bind Pods with Host Update Pods status Run Pods by kubelet

IBM’s contribution in Mesos Hierarchy Role (MESOS-6375) Multi-Role Frameworks (MESOS-1763) Scheduling (Pending) Resource Revocation Revocable resources by default Oversubscription for reservation Quota enhancements Optimistic Offers ….. Unified Container: Volume Unified Container: Network (CNI) Pods

(role = *, bigdata-daemons) Support Complex Business Container service BigData Service & Applications Hierarchy Role NS/Quota Spark Session Scheduler Myriad Masters Myriad Slaves (weight) Spark Slaves (weight) Dep 1 Dep 2 Dep 3 ns1 + quota1 ns2 + quota2 ns3 + quota3 Scheduling Kubernetes (role = *, bigdata-daemons) (role = bigdata-comute) Spark with kube-mesos Resource Requirement Mesos Resource Sharing App Area (label: app) BigData Area (label: bigdata) role = * Network/DNS Quota & Preemption role = bigdata-daemon: Reserve resources for HDFS and Yarn/Spark master role = bigdata-comute: Reserve resource for Yarn/Spark agents

Multi-Role & Hierarchy Role BigData ORG App others MR Team Spark Team Proj-1 Proj-2 50% Myriad BigData 40% Total Resources Spark 50% 60% Kubernetes NS-1 NS-2

Resources Requirements hostname=host1 hostname=host1 k8s Spark Myriad k8s Spark Myriad host ??? host1 host1 Mesos Mesos

Revocable Resources Support Mesos revocable resources KUbernetes Other Frameworks Support Mesos revocable resources Add a flag for revocable tasks/Pods Making scheduler to accept revocable offers Killing tasks when notified How to handle forceful eviction? Mesos Mesos Agent Mesos Agent …… Mesos Agent k8s-executor kubelet kubelet

What’s next?? … Support Sharing Policies & Smart Preemption: Revocable resources support (#19529) Scheduling enhancement (# 31068) Support Hierarchical Consumer: Namespace/Quota support/integrate (#31069) Multiple roles support Kube-DNS integrate with external DNS (# 28453) …

Roadmap of “Kuberntes on Mesos” v0.9 release new features v1.0 release Production Ready v0.8 release k8sm refactor v0.7 release new code base Nov, 2016 End of 2016 2017

Thank You !!