SlideShare una empresa de Scribd logo
1 de 44
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Auto Scaling Prime Time
Target Tracking Hits the Bullseye at
Tara Van Unen – Product Marketing, AWS
Vadim Filanovsky – Performance and Reliability Engineering, Netflix
C M P 3 1 1
N o v e m b e r 3 0 , 2 0 1 7
AWS re:INVENT
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Auto Scaling group
ELBCPU utilization
Amazon EC2
instances
Dynamic scaling
D e c r e a s e c o s t s
I m p r o v e a v a i l a b i l i t y
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
AWS customer survey 2016
…
Top Picks for Auto Scaling
MASTERING METRICS
MORE SERVICES
Speedy Set Up MORE SERVICES
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
• Multiple conditions
• Manual customization
• Granular control and tuning
Step scaling policies
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
T arg e t tracking
Set the target value on your
scaling metric and let target
tracking maintain your metric
T he rmostat
Set the temperature and let
the thermostat do the work
to maintain the temperature
Set it and forget it!
How it works
Cooling
74
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Speedy setup
Quick picks
• No conditions
• Self-optimizing
• Fastest scaling
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
ELB
Amazon EC2
instances
Traffic
5
10
15
20
25
30
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Instances
CPU
Target Utilization CPU Utilization Instances
Traffic
Traffic
Scaling-out
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Application Load Balancer
Auto Scaling group
Amazon EC2 instances
Availability Zone B
Load Balancing Layer
Compute
Layer
Data
Layer
Amazon
DynamoDB
…Demo
Amazon EC2 instances
Availability Zone A
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Cluster
Fleet
Aurora Replicas
EMR instance group
Scalable target Scalable dimension
Aurora
Service TasksECS
Spot Fleet request Spot InstancesEC2 Spot Fleet
AppStream 2.0
EMR
Core nodes
Task codes
AppStream Instances
Tables and global
secondary indices (GSI)
Provisioned capacityDynamoDB
More services
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Vancouver, BC – The home of Auto Scaling team
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Seattle, WA – Amazon headquarters
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Cluster footprintCluster footprint
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
ASG size ASG throughput
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
ASG size ASG throughput
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Chapter One
The need for Auto Scaling
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Encoding
556 Kbps 277 KbpsPerceptually optimal video encoding
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Encoding jobs run during off-peak
Normal usage
Encoding
3 am 7 pm
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Recommendations
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Regional failover
Other benefits of Auto Scaling
Red-black pushes Curing cancer
(Hack Day project)
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Auto
Scaling
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Chapter Two
Choosing the metric
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
What metric?
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
What metric, explained
Throughput per instance
Example: How much work I did
RequestCountPerTarget
Resource util. per instance
Example: How tired I am
avg. CPUUtilization
Pros: Direct measure of work; intuitive
Cons: Drifts over time
Pros: Requires less adjustment
Cons: More oscillation/jitter
VS
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Auto Scaling on multiple metrics
• Harder to reason about scaling behavior
• Different metrics might contradict each other, causing
oscillation
Typical Netflix setup
• Normal scale-up and -down on throughput
• Emergency scale-up on CPU (aka “the hammer rule”)
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Chapter Three
Setting the target
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
What is my target?
• curl, ab, siege, nghttp, Jmeter, Gatling…
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Squeezing with live prod traffic
Proxy ASG
Server ASG
Squeeze ASG
Clone
Client ASGs
Normal traffic flow
Controlled
throughput
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Understanding failures
VS
TODO - graph here
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Chapter Four
Scaling policy setup
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
What could go wrong?
70
90
110
Per instance throughput
Scale up
Scale down
0
100
200
300
400
0k
10k
20k
30k
40k
ASG size
ASG throughput
ASG
throughput
and size
Per-instance
throughput
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
“No rush” scaling
Problem: Scaling amounts
too small, cooldown too
long
Effect: Scaling lags behind
the traffic flow. Not
enough capacity at peak,
capacity wasted in trough.
Remedy: Increase scaling
amounts … or migrate to
target tracking!
0
100
200
300
400
15k
35k
55k
75k
ASG size
ASG throughput
40
80
120
160
Per instance
throughput
Scale up
Scale down
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Twitchy scaling
Problem: Scale-up policy is
too aggressive
Effect: Unnecessary
capacity churn
Remedy: Reduce scale-up
amount, increase the
number of eval periods …
or migrate to target
tracking!
0
25
50
75
100
0k
2k
3k
5k
6k
ASG size
ASG throughput
40
60
80
100
Per instance
throughput
Scale up
Scale down
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Should I stay or should I go?
Problem: -up and -down
thresholds are too close to
each other
Effect: Constant capacity
oscillation
Remedy: Move -up and -
down thresholds farther
apart … or migrate to
target tracking!
0
100
200
300
5k
15k
25k
35k
ASG size
ASG throughput
80
90
100
110
120
Per instance
throughput
Scale up
Scale down
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Step vs target tracking setup
VS
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Step vs target tracking setup
VS
and and
Scale-up
Scale-down
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Chapter Five
Traffic patterns
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Friday Saturday
What a difference
a day makes!
Understanding traffic
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Zooming in on morning traffic
2k
6k
10k
14k
18k
1:30 2:30 3:30 4:30 5:30 6:30 7:30
Weekday vs weekend traffic
Friday Saturday
ASG is at the lowest size
+
Aggressive traffic ramp-up
Can Auto Scaling keep up?
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
90
100
110
120
130
3:00 4:00 5:00 6:00 7:00
Per-instance throughput
30
60
90
120
150
3:00 4:00 5:00 6:00 7:00
ASG size
Too coldToo hot Saves from trouble Saves money
Target trackingStep vs
Target
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
AWS innovation Delivers Results
Target
tracking
Easier setup
Better demand
tracking
Per-second
billing
Hitting the bullseye
Productivity
Reliability
Cost
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
What are you scaling? #autoscaling
https://aws.amazon.com/autoscaling/getting-started
© 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Thank you!

Más contenido relacionado

La actualidad más candente

AWS Neptune - A Fast and reliable Graph Database Built for the Cloud
AWS Neptune - A Fast and reliable Graph Database Built for the CloudAWS Neptune - A Fast and reliable Graph Database Built for the Cloud
AWS Neptune - A Fast and reliable Graph Database Built for the CloudAmazon Web Services
 
Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...
Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...
Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...Amazon Web Services
 
갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...
갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...
갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...Amazon Web Services Korea
 
App Modernisation with Microsoft Azure
App Modernisation with Microsoft AzureApp Modernisation with Microsoft Azure
App Modernisation with Microsoft AzureAdam Stephensen
 
The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...
The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...
The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...Amazon Web Services
 
Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...
Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...
Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...Amazon Web Services
 
Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...
Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...
Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...Amazon Web Services
 
Introduction to AWS Lake Formation.pptx
Introduction to AWS Lake Formation.pptxIntroduction to AWS Lake Formation.pptx
Introduction to AWS Lake Formation.pptxSwathiPonugumati
 
엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...
엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...
엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...Amazon Web Services Korea
 
AWS Enterprise Day | Journey to the AWS Cloud
AWS Enterprise Day | Journey to the AWS CloudAWS Enterprise Day | Journey to the AWS Cloud
AWS Enterprise Day | Journey to the AWS CloudAmazon Web Services
 
AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...
AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...
AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...Amazon Web Services
 
Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...
Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...
Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...Amazon Web Services
 
Common Workloads on the AWS Cloud
Common Workloads on the AWS CloudCommon Workloads on the AWS Cloud
Common Workloads on the AWS CloudAmazon Web Services
 
Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...
Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...
Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...Amazon Web Services
 
[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager
[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager
[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial ManagerAmazon Web Services Korea
 
reInvent reCap 2022
reInvent reCap 2022reInvent reCap 2022
reInvent reCap 2022CloudHesive
 
Getting Started with Serverless Architectures
Getting Started with Serverless ArchitecturesGetting Started with Serverless Architectures
Getting Started with Serverless ArchitecturesAmazon Web Services
 

La actualidad más candente (20)

AWS Neptune - A Fast and reliable Graph Database Built for the Cloud
AWS Neptune - A Fast and reliable Graph Database Built for the CloudAWS Neptune - A Fast and reliable Graph Database Built for the Cloud
AWS Neptune - A Fast and reliable Graph Database Built for the Cloud
 
Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...
Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...
Tinder and DynamoDB: It's a Match! Massive Data Migration, Zero Down Time - D...
 
갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...
갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...
갤럭시 규모의 인공지능 서비스를 위한 AWS 데이터베이스 아키텍처 - 김상필 솔루션 아키텍트 매니저, AWS / 김정환 데브옵스 엔지니어,...
 
App Modernisation with Microsoft Azure
App Modernisation with Microsoft AzureApp Modernisation with Microsoft Azure
App Modernisation with Microsoft Azure
 
The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...
The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...
The Cloud Operating Model MVP: From Zero to Production Ready in 12 Weeks - Bu...
 
Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...
Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...
Behind the Scenes: Exploring the AWS Global Network (NET305) - AWS re:Invent ...
 
Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...
Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...
Moving Large Scale Contact Centers to Amazon Connect (BAP324) - AWS re:Invent...
 
Introduction to AWS Lake Formation.pptx
Introduction to AWS Lake Formation.pptxIntroduction to AWS Lake Formation.pptx
Introduction to AWS Lake Formation.pptx
 
Serverless Architectures.pdf
Serverless Architectures.pdfServerless Architectures.pdf
Serverless Architectures.pdf
 
엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...
엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...
엔터프라이즈의 효과적인 클라우드 도입을 위한 전략 및 적용 사례-신규진 프로페셔널 서비스 리드, AWS/고병률 데이터베이스 아키텍트, 삼성...
 
AWS Enterprise Day | Journey to the AWS Cloud
AWS Enterprise Day | Journey to the AWS CloudAWS Enterprise Day | Journey to the AWS Cloud
AWS Enterprise Day | Journey to the AWS Cloud
 
AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...
AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...
AWS re:Invent 2016: Building Enterprise Cloud Operations As a Service with T-...
 
Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...
Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...
Implementing Multi-Region AWS IoT, ft. Analog Devices (IOT401) - AWS re:Inven...
 
AWS-Data-Migration-module3
AWS-Data-Migration-module3AWS-Data-Migration-module3
AWS-Data-Migration-module3
 
Common Workloads on the AWS Cloud
Common Workloads on the AWS CloudCommon Workloads on the AWS Cloud
Common Workloads on the AWS Cloud
 
Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...
Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...
Visualize your data in Data Lake with AWS Athena and AWS Quicksight Hands-on ...
 
[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager
[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager
[AWS Builders 온라인 시리즈] AWS, 최적의 비용 효율화 방법은? - 조효원, AWS Cloud Financial Manager
 
reInvent reCap 2022
reInvent reCap 2022reInvent reCap 2022
reInvent reCap 2022
 
Getting Started with Serverless Architectures
Getting Started with Serverless ArchitecturesGetting Started with Serverless Architectures
Getting Started with Serverless Architectures
 
Big Data and Analytics on AWS
Big Data and Analytics on AWS Big Data and Analytics on AWS
Big Data and Analytics on AWS
 

Similar a Auto Scaling Prime Time: Target Tracking Hits the Bullseye at Netflix - CMP311 - re:Invent 2017

AWS reInvent 2017 recap - Optimizing Costs as You Scale on AWS
AWS reInvent 2017 recap - Optimizing Costs as You Scale on AWSAWS reInvent 2017 recap - Optimizing Costs as You Scale on AWS
AWS reInvent 2017 recap - Optimizing Costs as You Scale on AWSAmazon Web Services
 
Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...
Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...
Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...Amazon Web Services
 
How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...
How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...
How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...Amazon Web Services
 
Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...
Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...
Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...Amazon Web Services
 
Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017
Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017
Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017Amazon Web Services
 
透過Spot instances, Containers & Serverless降低成本
透過Spot instances, Containers & Serverless降低成本透過Spot instances, Containers & Serverless降低成本
透過Spot instances, Containers & Serverless降低成本Amazon Web Services
 
Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...
Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...
Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...Amazon Web Services
 
FSV305-Optimizing Payments Collections with Containers and Machine Learning
FSV305-Optimizing Payments Collections with Containers and Machine LearningFSV305-Optimizing Payments Collections with Containers and Machine Learning
FSV305-Optimizing Payments Collections with Containers and Machine LearningAmazon Web Services
 
Born in the Cloud, Built like a Startup
Born in the Cloud, Built like a StartupBorn in the Cloud, Built like a Startup
Born in the Cloud, Built like a StartupAmazon Web Services
 
CMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot Fleet
CMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot FleetCMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot Fleet
CMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot FleetAmazon Web Services
 
ARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost EfficiencyARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost EfficiencyAmazon Web Services
 
ARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost EfficiencyARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost EfficiencyAmazon Web Services
 
What's New in Serverless - SRV305 - re:Invent 2017
What's New in Serverless - SRV305 - re:Invent 2017What's New in Serverless - SRV305 - re:Invent 2017
What's New in Serverless - SRV305 - re:Invent 2017Amazon Web Services
 
CON320_Monitoring, Logging and Debugging Containerized Services
CON320_Monitoring, Logging and Debugging Containerized ServicesCON320_Monitoring, Logging and Debugging Containerized Services
CON320_Monitoring, Logging and Debugging Containerized ServicesAmazon Web Services
 
Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...
Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...
Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...Amazon Web Services
 
NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017
NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017
NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017Amazon Web Services
 
Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...
Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...
Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...Amazon Web Services
 
Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...
Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...
Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...Amazon Web Services
 
CMP314_Bringing Deep Learning to the Cloud with Amazon EC2
CMP314_Bringing Deep Learning to the Cloud with Amazon EC2CMP314_Bringing Deep Learning to the Cloud with Amazon EC2
CMP314_Bringing Deep Learning to the Cloud with Amazon EC2Amazon Web Services
 
DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...
DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...
DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...Amazon Web Services
 

Similar a Auto Scaling Prime Time: Target Tracking Hits the Bullseye at Netflix - CMP311 - re:Invent 2017 (20)

AWS reInvent 2017 recap - Optimizing Costs as You Scale on AWS
AWS reInvent 2017 recap - Optimizing Costs as You Scale on AWSAWS reInvent 2017 recap - Optimizing Costs as You Scale on AWS
AWS reInvent 2017 recap - Optimizing Costs as You Scale on AWS
 
Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...
Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...
Optimizing EC2 for Fun and Profit #bigsavings #newfeatures - CMP202 - re:Inve...
 
How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...
How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...
How Hess Has Continued to Optimize the AWS Cloud After Migrating - ENT218 - r...
 
Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...
Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...
Set it and Forget it: Auto Scaling Target Tracking Policies - AWS Online Tech...
 
Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017
Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017
Save up to 90% and Run Production Workloads on Spot - CMP307 - re:Invent 2017
 
透過Spot instances, Containers & Serverless降低成本
透過Spot instances, Containers & Serverless降低成本透過Spot instances, Containers & Serverless降低成本
透過Spot instances, Containers & Serverless降低成本
 
Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...
Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...
Case Study: Ola Cabs Uses Amazon EBS and Elastic Volumes to Maximize MySQL De...
 
FSV305-Optimizing Payments Collections with Containers and Machine Learning
FSV305-Optimizing Payments Collections with Containers and Machine LearningFSV305-Optimizing Payments Collections with Containers and Machine Learning
FSV305-Optimizing Payments Collections with Containers and Machine Learning
 
Born in the Cloud, Built like a Startup
Born in the Cloud, Built like a StartupBorn in the Cloud, Built like a Startup
Born in the Cloud, Built like a Startup
 
CMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot Fleet
CMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot FleetCMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot Fleet
CMP316_Hedge Your Own Funds Run Monte Carlo Simulations on EC2 Spot Fleet
 
ARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost EfficiencyARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost Efficiency
 
ARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost EfficiencyARC303_Running Lean Architectures How to Optimize for Cost Efficiency
ARC303_Running Lean Architectures How to Optimize for Cost Efficiency
 
What's New in Serverless - SRV305 - re:Invent 2017
What's New in Serverless - SRV305 - re:Invent 2017What's New in Serverless - SRV305 - re:Invent 2017
What's New in Serverless - SRV305 - re:Invent 2017
 
CON320_Monitoring, Logging and Debugging Containerized Services
CON320_Monitoring, Logging and Debugging Containerized ServicesCON320_Monitoring, Logging and Debugging Containerized Services
CON320_Monitoring, Logging and Debugging Containerized Services
 
Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...
Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...
Best practices for Running Spark jobs on Amazon EMR with Spot Instances | AWS...
 
NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017
NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017
NEW LAUNCH! Introducing AWS Fargate - CON214 - re:Invent 2017
 
Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...
Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...
Getting Started with Serverless Architectures with Microservices_AWSPSSummit_...
 
Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...
Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...
Auto Scaling: The Fleet Management Solution for Planet Earth - CMP201 - re:In...
 
CMP314_Bringing Deep Learning to the Cloud with Amazon EC2
CMP314_Bringing Deep Learning to the Cloud with Amazon EC2CMP314_Bringing Deep Learning to the Cloud with Amazon EC2
CMP314_Bringing Deep Learning to the Cloud with Amazon EC2
 
DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...
DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...
DynamoDB adaptive capacity: smooth performance for chaotic workloads - DAT327...
 

Más de Amazon Web Services

Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Amazon Web Services
 
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Amazon Web Services
 
Esegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateEsegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateAmazon Web Services
 
Costruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSCostruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSAmazon Web Services
 
Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Amazon Web Services
 
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Amazon Web Services
 
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...Amazon Web Services
 
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsMicrosoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsAmazon Web Services
 
Database Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareDatabase Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareAmazon Web Services
 
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSCrea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSAmazon Web Services
 
API moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAPI moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAmazon Web Services
 
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareDatabase Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareAmazon Web Services
 
Tools for building your MVP on AWS
Tools for building your MVP on AWSTools for building your MVP on AWS
Tools for building your MVP on AWSAmazon Web Services
 
How to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckHow to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckAmazon Web Services
 
Building a web application without servers
Building a web application without serversBuilding a web application without servers
Building a web application without serversAmazon Web Services
 
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...Amazon Web Services
 
Introduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceIntroduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceAmazon Web Services
 

Más de Amazon Web Services (20)

Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
 
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
 
Esegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateEsegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS Fargate
 
Costruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSCostruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWS
 
Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot
 
Open banking as a service
Open banking as a serviceOpen banking as a service
Open banking as a service
 
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
 
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
 
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsMicrosoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
 
Computer Vision con AWS
Computer Vision con AWSComputer Vision con AWS
Computer Vision con AWS
 
Database Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareDatabase Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatare
 
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSCrea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
 
API moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAPI moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e web
 
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareDatabase Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
 
Tools for building your MVP on AWS
Tools for building your MVP on AWSTools for building your MVP on AWS
Tools for building your MVP on AWS
 
How to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckHow to Build a Winning Pitch Deck
How to Build a Winning Pitch Deck
 
Building a web application without servers
Building a web application without serversBuilding a web application without servers
Building a web application without servers
 
Fundraising Essentials
Fundraising EssentialsFundraising Essentials
Fundraising Essentials
 
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
 
Introduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceIntroduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container Service
 

Auto Scaling Prime Time: Target Tracking Hits the Bullseye at Netflix - CMP311 - re:Invent 2017

  • 1. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Auto Scaling Prime Time Target Tracking Hits the Bullseye at Tara Van Unen – Product Marketing, AWS Vadim Filanovsky – Performance and Reliability Engineering, Netflix C M P 3 1 1 N o v e m b e r 3 0 , 2 0 1 7 AWS re:INVENT
  • 2. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Auto Scaling group ELBCPU utilization Amazon EC2 instances Dynamic scaling D e c r e a s e c o s t s I m p r o v e a v a i l a b i l i t y
  • 3. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. AWS customer survey 2016 … Top Picks for Auto Scaling MASTERING METRICS MORE SERVICES Speedy Set Up MORE SERVICES
  • 4. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. • Multiple conditions • Manual customization • Granular control and tuning Step scaling policies
  • 5. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
  • 6. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. T arg e t tracking Set the target value on your scaling metric and let target tracking maintain your metric T he rmostat Set the temperature and let the thermostat do the work to maintain the temperature Set it and forget it! How it works Cooling 74
  • 7. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Speedy setup Quick picks • No conditions • Self-optimizing • Fastest scaling
  • 8. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. ELB Amazon EC2 instances Traffic 5 10 15 20 25 30 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Instances CPU Target Utilization CPU Utilization Instances Traffic Traffic Scaling-out
  • 9. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Application Load Balancer Auto Scaling group Amazon EC2 instances Availability Zone B Load Balancing Layer Compute Layer Data Layer Amazon DynamoDB …Demo Amazon EC2 instances Availability Zone A
  • 10. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Cluster Fleet Aurora Replicas EMR instance group Scalable target Scalable dimension Aurora Service TasksECS Spot Fleet request Spot InstancesEC2 Spot Fleet AppStream 2.0 EMR Core nodes Task codes AppStream Instances Tables and global secondary indices (GSI) Provisioned capacityDynamoDB More services
  • 11. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
  • 12. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Vancouver, BC – The home of Auto Scaling team
  • 13. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Seattle, WA – Amazon headquarters
  • 14. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Cluster footprintCluster footprint
  • 15. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. ASG size ASG throughput
  • 16. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. ASG size ASG throughput
  • 17. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Chapter One The need for Auto Scaling
  • 18. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Encoding 556 Kbps 277 KbpsPerceptually optimal video encoding
  • 19. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Encoding jobs run during off-peak Normal usage Encoding 3 am 7 pm
  • 20. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Recommendations
  • 21. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Regional failover Other benefits of Auto Scaling Red-black pushes Curing cancer (Hack Day project)
  • 22. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Auto Scaling
  • 23. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Chapter Two Choosing the metric
  • 24. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. What metric?
  • 25. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. What metric, explained Throughput per instance Example: How much work I did RequestCountPerTarget Resource util. per instance Example: How tired I am avg. CPUUtilization Pros: Direct measure of work; intuitive Cons: Drifts over time Pros: Requires less adjustment Cons: More oscillation/jitter VS
  • 26. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Auto Scaling on multiple metrics • Harder to reason about scaling behavior • Different metrics might contradict each other, causing oscillation Typical Netflix setup • Normal scale-up and -down on throughput • Emergency scale-up on CPU (aka “the hammer rule”)
  • 27. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Chapter Three Setting the target
  • 28. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. What is my target? • curl, ab, siege, nghttp, Jmeter, Gatling…
  • 29. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Squeezing with live prod traffic Proxy ASG Server ASG Squeeze ASG Clone Client ASGs Normal traffic flow Controlled throughput
  • 30. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Understanding failures VS TODO - graph here
  • 31. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Chapter Four Scaling policy setup
  • 32. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. What could go wrong? 70 90 110 Per instance throughput Scale up Scale down 0 100 200 300 400 0k 10k 20k 30k 40k ASG size ASG throughput ASG throughput and size Per-instance throughput
  • 33. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. “No rush” scaling Problem: Scaling amounts too small, cooldown too long Effect: Scaling lags behind the traffic flow. Not enough capacity at peak, capacity wasted in trough. Remedy: Increase scaling amounts … or migrate to target tracking! 0 100 200 300 400 15k 35k 55k 75k ASG size ASG throughput 40 80 120 160 Per instance throughput Scale up Scale down
  • 34. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Twitchy scaling Problem: Scale-up policy is too aggressive Effect: Unnecessary capacity churn Remedy: Reduce scale-up amount, increase the number of eval periods … or migrate to target tracking! 0 25 50 75 100 0k 2k 3k 5k 6k ASG size ASG throughput 40 60 80 100 Per instance throughput Scale up Scale down
  • 35. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Should I stay or should I go? Problem: -up and -down thresholds are too close to each other Effect: Constant capacity oscillation Remedy: Move -up and - down thresholds farther apart … or migrate to target tracking! 0 100 200 300 5k 15k 25k 35k ASG size ASG throughput 80 90 100 110 120 Per instance throughput Scale up Scale down
  • 36. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Step vs target tracking setup VS
  • 37. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Step vs target tracking setup VS and and Scale-up Scale-down
  • 38. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Chapter Five Traffic patterns
  • 39. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Friday Saturday What a difference a day makes! Understanding traffic
  • 40. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Zooming in on morning traffic 2k 6k 10k 14k 18k 1:30 2:30 3:30 4:30 5:30 6:30 7:30 Weekday vs weekend traffic Friday Saturday ASG is at the lowest size + Aggressive traffic ramp-up Can Auto Scaling keep up?
  • 41. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. 90 100 110 120 130 3:00 4:00 5:00 6:00 7:00 Per-instance throughput 30 60 90 120 150 3:00 4:00 5:00 6:00 7:00 ASG size Too coldToo hot Saves from trouble Saves money Target trackingStep vs Target
  • 42. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. AWS innovation Delivers Results Target tracking Easier setup Better demand tracking Per-second billing Hitting the bullseye Productivity Reliability Cost
  • 43. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. What are you scaling? #autoscaling https://aws.amazon.com/autoscaling/getting-started
  • 44. © 2017, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Thank you!