Getting Started with Amazon EC2 and Compute Services

© 2016, Amazon Web Services, Inc. or its Affiliates. All rights reserved.

Sebastian Dreisch, Global Bus Dev – Compute Services

July 13, 2016

Getting Started with Amazon EC2

and AWS Compute Services

AWS compute offerings

LambdaServerless compute

platform for stateless

code execution in

response to triggers

ECSContainer

management service

for running Docker on

a managed cluster of

EC2

EC2Virtual servers

in the cloud

EC2 – Virtual servers in the cloud?

AWS global infrastructure

Over 1 million active customers across 190 countries

2,300 government agencies

7,000 educational institutions

22,000 nonprofits

13 regions

35 Availability Zones

*9 more Availability Zones and 4 more regions coming online this year

Amazon Elastic Compute Cloud (EC2) -

Elastic virtual servers in the cloud

Physical servers in AWS global regions

Host server

Hypervisor

Guest 1 Guest 2 Guest n

Amazon EC2 ten years ago…

First generation, single instance family and size

• m1.small (1 vCPU, 1.7 GiB RAM, 160 GB storage)

Linux only

On-Demand pricing only

EC2 instances today

c4.largeInstance family

Instance generation

Instance size

Performance factor: CPU

Intel Xeon E5-2670 (Sandy Bridge) CPUs

• Available on M3, CC2, CR1, and G2 instance types

Intel Xeon E5-2680 v2 (Ivy Bridge) CPUs

• Available on C3, R3, and I2 instance types

• 2.8 GHz in C3, Turbo enabled up to 3.6 GHz

• Supports enhanced Advanced Vector Extensions (AVX) instructions

Intel Xeon E5-2666 v3 (Haswell – AVX2) CPUs

• Available on C4, D2, and M4 instance types

• 2.9 GHz in C4, Turbo enabled up to 3.5 GHz (with Intel Turbo Boost)

• Supports AVX2 instructions

http://aws.amazon.com/ec2/instance-types/

Performance factor: NetworkingDevice passthrough: Enhanced networking

• SR-IOV eliminates need for driver domain

• Physical network device exposes virtual function to instance

• Enhanced Networking is currently supported in R3, C3, C4, M4, D2, and I2 instances

enables you to get significantly higher packet per second (PPS) performance, lower network jitter, and lower latencies

uses a new network virtualization stack that provides higher I/O performance and lower CPU utilization compared to

traditional implementations

New: Elastic network adapter available now for the new X1 instance type!

• Next generation of Enhanced Networking

• Hardware checksums

• Multi-queue support

• Receive-side steering

• 20 Gbps in a placement group

delivering high throughput and great packet per second (PPS) performance, minimizes the load on the host processor in a

number of ways, and also does a better job of distributing the packet processing workload across multiple vCPUs

1 2 4 8 16 321

2

4

8

16

32

64

128

256

Me

mo

ry (

GB

)

vCPU

g2.2xlarge8 vCPU, 15 GB1 x 60 SSDNVIDIA GPU (1,536 CUDA cores, 4GB Mem)

4 vCPU, 30.5 GBi2.xlarge (High IO) - 1 x 800 SSDd2.xlarge (Dense) - 3 x 2000 HDD

8 vCPU, 61 GBi2.2xlarge (High IO) - 2x800 SSDd2.2xlarge (Dense) - 6 x 2000 HDD

16 vCPU, 122 GBi2.4xlarge (High IO) - 4x800 SSDd2.4xlarge (Dense) - 12x2000 HDD

32 vCPU, 244 GBi2.8xlarge (High IO) - 8x800 SSD

36 vCPU, 244 GBd2.8xlarge (Dense) - 24x2000 HDD

m3.xlarge4 vCPU, 15 GB2 x 40 SSD

m3.2xlarge8 vCPU, 30 GB2 x 80 SSD

m3.large2 vCPU, 7.5 GB1 x 32 SSDm3.medium

1 vCPU, 3.75 GB, 1 x 4 SSD

t2.micro1 vCPU, 1GBEBS Only

t2.small1 vCPU, 2GBEBS Only

t2.medium2 vCPU, 4GBEBS Only

r3.large2 vCPU, 15.25 GB1 x 32 SSD

r3.xlarge4 vCPU, 30.5 GB1 x 80 SSD

r3.2xlarge8 vCPU, 61 GB1 x 160 SSD



2 vCPU, 3.75 GBc4.large - EBS Onlyc3.large - 2 x 16 SSD

4 vCPU, 7.5 GBc4.xlarge - EBS Onlyc3.xlarge - 2 x 40 SSD

8 vCPU, 15 GBc4.2xlarge - EBS Onlyc3.2xlarge - 2 x 80 SSD


m4.large2 vCPU, 8 GBEBS Only

m4.xlarge4 vCPU, 16 GBEBS Only

m4.2xlarge8 vCPU, 32 GBEBS Only

m4.4xlarge16 vCPU, 64 GBEBS Only

m4.10xlarge40 vCPU, 160GBEBS Only

t2.large2 vCPU, 8 GBEBS Only

Storage Optimized

GPU Instances

General Purpose

Memory Optimized

Compute Optimized

New M4s/T2 Large

t2.nano1 vCPU, 512MBEBS Only

g2.8xlarge32vCPU, 60 GB2 x 120 SSD4 NVIDIA GPUs (1,536 CUDA cores, 4GB Mem)


39 (latest generations) EC2 instance types

Performance factor: Memory

Performance factor: Storage (Options)

Locally attached or “instance storage”

Network attached:

Amazon EBS General Purpose (SSD) volumes

Amazon EBS Provisioned IOPS (SSD) volumes

Amazon EBS Magnetic volumes (multiple types for different use cases)

Amazon EFS (seconds to create a scalable shared NFSv4 file system)

Amazon S3 and Amazon Glacier for object storage

EC2 – Why use servers in the cloud?

Fast deploymentsAccess computing

infrastructure in minutes

Low costPay-as-you-go pricing

ElasticEasily add or remove capacity

Globally accessibleEasily support customers

around the world

SecureA collection of tools to

protect data and privacy

ScalableAccess to effectively

limitless capacity

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day

Capacity of 1 server

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day


Traditional capacity required

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day



1 server for 8 hours

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day



1 server for 8 hours 1 server for 8 hours

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day





0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day






0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Serv

er

load

Hour of day



1/3rd

saving

0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Insta

nce c

ou

nt

Day of month

0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Insta

nce c

ou

nt

Day of month

Monthly

predictable

peak

processing

0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Insta

nce c

ou

nt

Day of month


0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Insta

nce c

ou

nt

Day of month

Elastic capacity


0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Insta

nce c

ou

nt

Day of month

75% savings


Elastic capacity

Tooling - Scale automatically

Amazon EC2 Auto Scaling ELB

Actual

EC2

Elastic virtual servers

in the cloud

Dynamic traffic

distribution

Automated scaling

of EC2 capacity

2007 2008 2009 2010 2011 2012 2013 2014 2015

9 24 48 61 82159

280

514

722

Rapid pace of customer-driven improvements

There is no compression algorithm for experience

AWS feature and service launches(Above & beyond all the regular updates to the infrastructure platform)

Rapid pace of customer driven improvements

Security - Our top priority!

AWS feature and service launches(Above and beyond the regular updates to the infrastructure platform)

2007 2008 2009 2010 2011 2012 2013 2014

9 24 48 61 82159

280

514

Security, compliance, governance,and/or audit capabilities

Consistent, regular, exhaustive third-party evaluations

• Secured premises

• Secured access

• Built-in firewalls

• Unique users

• Multi-factor authentication

• Private subnets

• Encrypted data storage

• Dedicated connection

Architected for enterprise security

Access a deep set of cloud security tools

Encryption

AWS Key

Management

Service

CloudHSM Server-side

encryption

Networking

Virtual

Private

Cloud

Web

application

firewall

Compliance

ConfigCloudTrailAWS Service

Catalog

Identity

IAM Active

Directory

integration

SAML

federation

EC2 – How do I get started?

http://aws.amazon.com/getting-started/

Just get started - for free! (AWS Console)

Use the AWS Marketplace

Browse, search, discover, and launch thousand of AWS Marketplace Amazon Machine Images (AMIs) directly from the Amazon EC2 console

2,700+ products listed in 35 categories

software listings from more than 925 ISVs

Build reliable architectures

Easily build highly available applications

ELB distributes load

Auto Scaling helps ensure availability and scale

Use multiple Availability Zones (AZs)

Use multiple global regions

Example: 3-tier web application architecture

Build secure architectures

Use VPC - Provision a logically isolated section of the AWS cloud

Control your virtual networking environment with:• Subnets

• Route tables

• Security groups

• Network ACLs

• Flow logging (new!)

Control if and how your instances access the Internet

Connect to your on-premises network through a hardware VPN or AWS Direct Connect

Monitor all changes through Amazon CloudWatch Logs and AWS CloudTrail

Availability Zone 1a Availability Zone 1b

Internet

10.0.0.5

10.0.0.6

10.0.3.17

10.0.3.5

10.0.1.5

10.0.1.25

10.0.1.8

10.0.1.6

VPC subnet

VPC subnet

VPC subnet

Virtual private gateway

Customer gateway

VPN connection

Internet gateway

Customer data center

Example: Enterprise application architecture

On-Demand

Pay for compute

capacity by the

hour with no long-

term commitments

For spiky

workloads or to

define needs

Reserved

Make a low, one-

time payment and

receive a

significant discount

on the hourly

charge

For committed

utilization

Spot

Bid for unused

capacity, charged at

a Spot Price that

fluctuates based on

supply and demand

For time-insensitive

or transient

workloads

Dedicated

Launch instances in

Amazon VPC that

run on hardware

dedicated to a single

customer

For BYOL and highly

sensitive/regulated

workloads

Use a purchasing option (mix) that best fits your workload

RIs for cost savings and reserved capacity

The No Upfront Reserved Instance option provides a discount compared to On-Demand (up to 55% off) and does not require an upfront payment.

The Partial Upfront option balances the payments of an RI between upfront and hourly and provides a higher discount (up to 76%) compared to the No Upfront option.

With the All Upfront option, you benefit from the highest discount compared to On-Demand (up to 77% off).

$-

$500

$1,000

$1,500

$2,000

$2,500

$3,000

30% 40% 50% 60% 70% 80% 90% 100%

Utilization

m3.xlarge 1yr OD/RI Break Even Utilization

On Demand

No Upfront

Partial Upfront

All Upfront

Spot for interruptible workloads and best pricing

Best Spot use cases include any batch-oriented, fault-tolerant application

What have customers done on EC2?

18 hours

205,000 materials analyzed

156,314 AWS Spot cores at peak

2.3M core-hours

Total spending: $33K

(Under 1.5 cents per core-hour)

Use dedicated hosts to enable BYOL

Host ID = h-123abc

Sockets = 2

Physical cores = 20

• Granular resource and placement controls• Dedicated Host allocation

• Granular instance placement

• Instance-host affinity

• Visibility into physical resources • Physical core and socket counts

• Capacity utilization

• Instance location

Dedicated Host configurations

A C4 Dedicated Host provides capacity for 8 c4.xlarge

instances. Every C4 dedicated host is supported by 2

sockets and 20 physical cores.

Dedicated Host attributes # of instances per host by instance size

Instance family Sockets

Physical

cores medium large xlarge 2xlarge 4xlarge 8xlarge 10xlarge

c3 2 20 - 16 8 4 2 1 -

c4 2 20 - 16 8 4 2 1 -

g2 2 20 - - - 4 - 1 -

m3 2 20 32 16 8 4 - - -

d2 2 24 - - 8 4 2 1 -

r3 2 20 - 16 8 4 2 1 -

m4 2 24 - 22 11 5 2 - 1

i2 2 20 - - 8 4 2 1 -

ECS – Why use it?

Amazon EC2 Container Service is a highly scalable, high-

performance container management service that supports

Docker containers and allows you to easily run applications

on a managed cluster of Amazon EC2 instances.

Amazon ECS

Docker

Task

Container instance

Amazon

ECS

Container

ECS Agent

ELB

Internet

ELB

User/

scheduler

API

Cluster management engine

Task

Container

Docker

Task

Container instance

Container

ECS Agent

Task

Container

Docker

Task

Container instance

Container

ECS Agent

Task

Container

AZ 1 AZ 2

Key/value store

Agent communication service

It’s easy and FREE!

• Please visit:

https://aws.amazon.com/ecs/getting-started/

Lambda – Serverless code execution?

High performance at any scale;

cost-effective and efficient

No infrastructure to manage

Pay only for what you use: Lambda

automatically matches capacity to

your request rate. Purchase

compute in 100 ms increments.

Bring your own code

Stateless, trigger-based code execution

Run code in a choice of standard

languages. Use threads, processes,

files, and shell scripts normally.

Focus on business logic, not

infrastructure. You upload code; AWS

Lambda handles everything else.

AWS Lambda functions

Lambda – Why use it?

No server is easier to manage

than no server

Lambda – How do I use it?

Key Lambda scenarios

Data processing

Stateless processing of

discrete or streaming

updates to your data

store or message bus

Control systems

Customize responses

and response workflows

to state and data

changes in AWS

App back-end

development

Execute server-side

back-end logic in a

cross-platform fashion

AWS Lambda use case – Data processing

Please tell us about what you are building next!

Getting Started with Amazon EC2 and Compute Services

Technology