Readings

• A quick introduction to Docker

http://blog.scottlowe.org/2014/03/11/a-quick-introduction-t
o-docker/

• Docker Get Started, Part 1: Orientation and setup

https://docs.docker.com/get-started/

• Microservices
https://martinfowler.com/articles/microservices.html
 The Challenge
User DB

Do services and
appropriately?
apps interact
Static website postgresql + pgv8 + v8
Multiplicity of

Queue Analytics DB
nginx 1.5 + modsecurity + openssl
hadoop + hive + thrift + OpenJDK
Stacks

+ bootstrap 2 Redis + redis-sentinel

Web frontend
Background workers
Ruby + Rails + sass + Unicorn
Python 3.0 + celery + pyredis + libcurl +
ffmpeg + libopencv + nodejs + phantomjs API endpoint
Python 2.7 + Flask + pyredis + celery +
psycopg + postgresql-client

Production Cluster
environments

Development VM
Multiplicity of

Public Cloud

smoothly and
Can I migrate
hardware

quickly?
QA server
Disaster recovery
Customer Data Center Contributor’s laptop
Production Servers
Results in N X N compatibility nightmare
Static website ? ? ? ? ? ? ?

Web frontend ? ? ? ? ? ? ?

Background ? ? ? ? ? ? ?
workers

User DB ? ? ? ? ? ? ?

Analytics DB ? ? ? ? ? ? ?

Queue ? ? ? ? ? ? ?

Develop QA Single Onsite Public Contribu Custome

ment Server Prod Cluster Cloud tor’s r
VM Server laptop Servers
 Do I worry about Can I transport
how goods quickly and
interact (e.g. smoothly
coffee beans next (e.g. from boat to
to spices) train to truck)
Cargo Transport Pre-1960 ng
transporting/stori
Goods methods for
Multiplicity of Multipilicity of
NxN Matrix?
? ? ? ? ? ? ?

? ? ? ? ? ? ?

? ? ? ? ? ? ?

? ? ? ? ? ? ?

? ? ? ? ? ? ?

? ? ? ? ? ? ?
 Solution: Shipping Container

Do I worry about how

coffee beans next to
goods interact (e.g.
Multiplicity of Goods

spices)
A standard container
that is loaded with
virtually any goods,
and stays sealed until
it reaches final
delivery.
…in between, can be
loaded and unloaded,

(e.g. from boat to train

Multiplicity of methods

stacked, transported

quickly and smoothly

transporting/storing

efficiently over long

Can I transport
distances, and

to truck)
transferred from one
for

mode of transport to
another
This eliminated the NXN problem…
• 90% of all cargo now shipped in a standard container
• Order of magnitude reduction in cost and time to load and
unload ships
• Massive reduction in losses due to theft or damage
• Huge reduction in freight cost as percent of final goods (from
>25% to <3%)
 massive globalizations
• 5000 ships deliver 200M containers per year
 Docker is a shipping container system for code

Do services and apps

Static website User DB Web frontend Queue
Multiplicity of Stacks

Analytics DB

appropriately?
interact
An engine that
enables any payload
to be encapsulated
as a lightweight,
portable, self-
sufficient
container…
…that can be
manipulated using

smoothly and quickly

environments
Multiplicity of

standard operations and

hardware

Can I migrate
run consistently on
virtually any hardware
platform

Develop QA server Customer Public Cloud Production Contributo

ment VM Data Center Cluster r’s laptop
 Or…put more simply

Do services and apps

Static website User DB Web frontend Queue Analytics DB
Multiplicity of Stacks

appropriately?
interact
Developer: Build
Once, Run Anywhere
(Finally)

Operator: Configure
Once, Run Anything

smoothly and quickly

environments
Multiplicity of
hardware

Can I migrate
Develop QA server Customer Public Cloud Production Contributo
ment VM Data Center Cluster r’s laptop
Docker solves the NXN problem
Static website

Web frontend

Background
workers

User DB

Analytics DB

Queue

Develop QA Single Onsite Public Contribu Custome

ment Server Prod Cluster Cloud tor’s r
VM Server laptop Servers
Container
● Container is a type of virtualization that allows multiple isolated
applications to run on a single host operating system.

● Containers are lightweight and portable, and they provide a way to

package an application and its dependencies together, so it can run
consistently across different environments.

● Containers are different from virtual machines (VMs) as they do not

require a separate operating system to be installed for each
container. This avoids the overhead of running multiple separate
operating system.
Container (cont)
● The core component of container virtualization is the container
engine, which is a software that manages the creation and
execution of containers.
● The container engine uses the host operating system's kernel to
create isolated environments for each container.
● These isolated environments, called containers which share the host
operating system's kernel and libraries, but have their own file
system, network stack, and process space.
● Containers are created from container images, which are pre-
configured and bundled with all the necessary dependencies for the
application to run.
● Container images can be stored in a container registry, and can be
easily pulled and run on any host that supports the container
engine.
LXC
● In Linux, the most common container technology is LXC
(Linux Containers).
● LXC uses Linux kernel features such as control groups
and namespaces to create isolated environments for
containers.
● Control groups, also known as cgroups, are used to limit,
prioritize and account system resources such as CPU,
memory, and I/O.
● Namespaces, on the other hand, are used to provide
isolated environments for container's process, network
and file system.
Virtual Machines vs Containers
Virtual Machines vs Containers (Cont.)
• Containers share resources with the host OS, which
makes them an order of magnitude more efficient

• Applications running in containers incur little to no

overhead compared to applications running natively on
the host OS.

• The fundamental goals of VMs and containers are

different
• the purpose of a VM is to fully emulate a foreign environment
• the purpose of a container is to make applications portable and
self-contained
Containers over VMs
1. Cloud-native applications: Containerized applications are designed to
run in distributed environments and can be easily deployed to different
cloud platforms, making them well suited for cloud-native applications.
2. Microservices: Containers are well suited for microservices
architecture, as they provide isolation and are lightweight, making it
easy to deploy and scale individual services.
3. Resource efficiency: Containers are more lightweight and efficient than
VMs, as they share the host operating system's kernel and libraries,
which reduces the resources required to run the application.
4. Rapid Development and Testing: Containers allow developers to
quickly create and test new applications, as they can be easily created
and destroyed, which can help to speed up the development process.
VMs over Containers
1. Legacy Applications: If an application is not designed to run in a containerized
environment, it may be more efficient to run it in a virtual machine, as it will not
require any modification.
2. High Security Requirements: In some cases, such as for sensitive data or applications
with strict regulatory compliance requirements, virtual machines may be a better
option as they provide a higher level of isolation and security than containers.
3. Hardware-intensive Workloads: Applications that require direct access to physical
hardware resources such as GPUs, high-speed networking, or specialized devices
may be better suited for virtual machines as they can provide direct access to these
resources.
4. Different Operating Systems: If your application requires different operating systems
for different components, it may be more efficient to use VMs, as each VM can run a
different operating system.
5. High Performance Computing: Applications that need high-performance computing
resources like large memory, high-speed storage, or complex network configurations
are better suited for virtual machines.
Docker
● Docker is an open-source platform that is widely used
container engine on Linux.
● Docker provides a simple and efficient way to create,
deploy, and run containerized applications.
● It also provides an easy-to-use command-line interface
and an API that allows you to automate the
management of your containers.
Docket Engine Overview
Container Orchestration
● Container orchestration is the process of automating the deployment, scaling, and
management of containerized applications.
● It involves the use of software tools that can manage and coordinate the scheduling, scaling,
and deployment of containers across a cluster of machines.
● With container orchestration, administrators can define the desired state of their applications,
such as the number of replicas of a container, and the orchestration tool will automatically
ensure that the desired state is met by creating, updating, or deleting containers as needed.

● Some of the key features of container orchestration include:

○ Automatic scaling: The ability to automatically scale the number of containers in response
to changes in demand.

○ Self-healing: The ability to automatically recover from failures, such as if a container

crashes, by restarting it or rescheduling it on another node.

○ Load balancing: The ability to automatically distribute incoming traffic across multiple
containers.

○ Service discovery: The ability to automatically discover and connect to other services in
the system.
Docker Swarm
● Docker Swarm is a native clustering and orchestration
solution for Docker.
● It allows you to create and manage a cluster of Docker
nodes as a single virtual system.
● It provides features such as service discovery, load
balancing, and scaling, which makes it easy to build,
ship and run distributed applications.
● It also allows you to schedule the placement of
containers across the nodes in the cluster, and to ensure
that the desired number of replicas of a container are
running at all times.
Kubernetes
● Kubernetes (often shortened to "K8s") is an open-source container orchestration
system that automates the deployment, scaling, and management of containerized
applications.
● It was originally developed by Google and is now maintained by the Cloud Native
Computing Foundation (CNCF).
● Kubernetes manage containerized applications in a clustered environment. It
abstracts away the underlying infrastructure and provides a consistent set of APIs
for deploying, scaling, and managing containerized applications.
● Kubernetes uses a declarative configuration model, where the desired state of the
system is defined in configuration files.
● The declarative configuration model is a way of describing the desired state of a
system, rather than the steps needed to achieve that state.
● This approach allows administrators to specify the desired state of their applications
and infrastructure, and then rely on Kubernetes to automatically make the
necessary changes to achieve that state.
Kubernetes (Cont.)
•Kubernetes provides several key features:

● Automatic scaling of application replicas based on resource

usage
● Self-healing capabilities to automatically replace and
reschedule failed containers
● Automated rollouts and rollbacks of application updates
● Service discovery and load balancing for application
components
● Automated storage provisioning and management
● Automatic binpacking of containers onto nodes in the cluster
Docker Swarm Vs Kubernetes
•Kubernetes and Docker Swarm are both popular container orchestration systems. However, they have some
key differences:

1. Kubernetes is an open-source platform that was originally developed by Google and is now maintained by
the Cloud Native Computing Foundation (CNCF). Docker Swarm, on the other hand, is a native clustering
and orchestration solution for Docker.
2. Kubernetes provides a more powerful and flexible orchestration engine, with more advanced features such
as automatic scaling, self-healing, and rollouts, and rollbacks. Docker Swarm is more limited in terms of its
orchestration capabilities, but it is simpler to set up and use.

3. Kubernetes provides a declarative configuration model where the desired state of the system is defined in
configuration files, while Docker Swarm uses a more imperative model where the desired state is defined
through commands.

4. Kubernetes supports a wide range of deployment options including on-premise, in the public cloud, and
hybrid deployments. Docker Swarm is typically used for deployments on a single cluster or a single cloud
provider.

5. Kubernetes has a large and active community, with many contributors and a wide range of third-party tools
and plugins available. Docker Swarm has a smaller community and fewer third-party tools and plugins.

6. Kubernetes is more complex to set up and manage, it is typically used for large-scale production
deployments, while Docker Swarm is simpler and more suited for small-scale and development
deployments.
Microservice Architecture
“Microservice architectural is an approach to developing a
single application as a suite of small services, each
running in its own process and communicating with
lightweight mechanisms, often an HTTP resource API.
These services are built around business capabilities and
independently deployable by fully automated deployment
machinery.”
Microservice Architecture (cont.)
Traditional development

Microservices Architecture
Coming up next!
Working with Docker . . .