Downloaded 18 times
Uploaded byAmin Afshar
Containers and Docker
This document discusses containers and Docker. It provides an overview of containers compared to virtual machines and how they leverage features of the Linux kernel like namespaces and cgroups. It describes the history and components of Docker, including images, Dockerfile, registry, and engine. Benefits of Docker for development, deployment, and portability are outlined. Docker Compose is introduced for multi-container applications. Finally, using GPUs with Docker and Nvidia is briefly covered.
More Related Content
Similar to Containers and Docker
Containers and Docker
- 1. CONTAINERS AND DOCKER AminAfshar System Administrator Avid Arvand Shahid Beheshti University Natural Language Processing (NLP) Lab May 2019
- 2. AGENDA Virtualization /Hypervisors / Virtual Machines Containers Containers vs. Virtual Machines Linux kernel features Docker Platform Demo
- 3.
- 4. HYPERVISORS BARE METAL, NATIVEOR TYPE I
- 5.
- 6.
- 7. CONTAINERS Containers areprocesses Containers are isolated processes Share the same OS kernel OS-level virtualization Another abstraction layer on top of the hypervisor
- 8.
- 9.
- 10. CONTAINERS VS. VIRTUALMACHINES
- 11. CONTAINERS AND VIRTUALMACHINES
- 12. LINUX KERNEL FEATURES: NAMESPACES Controlwhat a process can see. Namespaces “trick” a process into seeing an entirely separate machine. originated in 2002 in the 2.4.19 kernel
- 13.
- 14.
- 15. LINUX KERNEL FEATURES: CGROUPS Control what a process can use. Control Groups (cgroups) isolate the resource (cpu, memory, …) usage. Started at Google in 2006 Merged into linux kernel mainline v2.6.24 released in Jan. 2008. Redesign in 2013 in kernel v3.15
- 16.
- 17.
- 19.
- 20. MOST POPULAR ANDWIDELY USED CONTAINER PLATFORM Solomon Hykes started Docker as an internal project within dotCloud PaaS PyCon 2013: The Future of Containers Open sourced in March 2013 Introduced an ecosystem for managing containers Incredible adoption rate in just few years A highly efficient, layered container image model A global and local container registries A clean REST API, and a CLI
- 21. Docker Platform Image The basisof a Docker container. Container The image when it is ‘running.’ Docker Engine Manages Images and Containers. Registry Stores, distributes and manages Docker images
- 22.
- 23.
- 24. Docker Image • aread-only template • Built up from a series of layers • Each layer represents an instruction in the image’s Dockerfile • Each layer except the very last one is read-only
- 25. Dockerfile FROM nginx ENV AUTHOR=Docker WORKDIR/usr/share/nginx/html COPY Hello_docker.html /usr/share/nginx/html CMD cd /usr/share/nginx/html && sed -e s/Docker/"$AUTHOR"/ Hello_docker.html > index.html ; nginx -g 'daemon off;'
- 26.
- 27.
- 28.
- 29. WHY DOCKER TheIT landscape is changing Microservices DevOpsCloud Migrate workloads to cloud Portability across environments
- 30. WHY DOCKER Applicationsare transforming Loosely Coupled Services Many Small Servers or devices ~2000 Today Monolithic Big Servers Slow changing Rapidly updated
- 31.
- 32. WHY DOCKER Tugof War Between Developers and Ops (Diverse Organizations) Developers IT Operations
- 33. WHY DOCKER MultipleStacks, Multiple Stages (CI/CD) = Complexity 1. Development 2. Test 3. Stage / Production
- 34. WHY DOCKER SOLVINGTHE DEPLOYMENT MATRIX
- 38. Build, Ship, Run BUILD(Development) Development Environments SHIP (Distribution) Secure Content & Collaboration RUN (Deployment) Deploy, Manage, Scale Registry Multi-container apps Engines running on servers in cloud or datacenter Images stored in repos Clients pull and push images
- 39. BENEFITS OF DOCKER Standardized packaging format for software and dependencies (primary and biggest benefit) Portability, ability to move between infrastructure Dev/Prod Parity Increase Agility Speed, No OS to boot, applications online in seconds Efficiency, Less OS overhead, ImprovedVM density Natural fit for microservices style apps (self-contained services) Secure, reduced attack surface
- 41. • A toolfor defining and running multi-container Docker applications • Use a Compose file to configure application’s services • With a single command, create and start all the services from configuration (docker- compose up) • Single host deployments • Great for development, testing, and staging environments, as well as CI workflows • Speed up project onboarding Docker Compose
- 42. docker-compose.yml version: '3' services: web: build: . ports: -"5000:5000" volumes: - .:/code - logvolume01:/var/log links: - redis redis: image: redis volumes: logvolume01: {}
- 43.
- 45. • Configure gpuaccess for container leveraging nvidia hardware • Linux kernel > 3.10 nvidia architecture > fermi 2.1 2011 • Install cuda-drivers • Install nvidia-docker2 and restart docker daemon • Use nvidia/cuda images Docker and nvidia
- 46.
- 47.