SlideShare a Scribd company logo
메가존과 AWS가 공개하는 AWS 비용 최적화 전략
김성용 매니져, MegaZone
이우상 매니져, AWS
• Basics
• Structure your Strategy
• Conclusion
Before Basics..
On and Off Fast Growth
Predictable peaksVariable peaks
WASTE (비용 낭비)
CUSTOMER DISSATISFACTION (고객 이탈 발생)
Before Basics..
Amazon EBS standa
rd volume prices are
lowered up to 50%
for both storage and
I/O requests.
January 21, 2014
50%
Amazon is reducing
prices for Amazon
ElasticCache cache
nodes by an average
of 34%.
March 26, 2014
34%
Amazon S3: We are
reducing prices for
Standard and Reduced
Redundancy Storage, by
an average of 51%.
March 26, 2014
51%
1. Do Nothing
50 times since our inception in 2006. Recent price drops included…
Before Basics..
2. Do Almost Nothing
CloudWatch, Trusted Advisor, Start Small
Before Basics..
3. Do Iterating Architecture Optimization
Basics
4. 7 tips for Architecture Optimization
• Turn off when not using (working 12 hours vs 24 hours)
• Use Auto Scaling
• Use Reserved Instance (1Y 35% ↓, 3Y 75% ↓)
• Use Spot Instance (up to 90% ↓)
• Use S3 options (S3, IA, Glacier → 3¢, 1.3 ¢, 0.7 ¢)*
• Use AWS Dynamo DB (dynamic provisioning)
• Remove un-needed component and utilize
AWS managed services
* Based on N. Virginia Region, for more details visit : https://aws.amazon.com/s3/pricing/?nc1=h_ls
• Basics
• Structure your Strategy
• Conclusion
Structure your Strategy
1. Have visibility on your cost
2. Optimize AWS usage
3. Use Reserved Instance
Structure your Strategy
1. Have visibility on your cost
1. Budgets
2. Cost Explorer
3. Cost and Usage Report
2. Optimize AWS usage
1. Trusted Advisor
2. Right Sizing
3. Elasticity
3. Use Reserved Instance
Have visibility on your cost
AWS BUDGETS
• Define and track budgets for your AWS costs,
forecast your AWS costs
• Budgets takes place on a fine-grained basis,
with filtering or customization based on Availability Zone,
Linked Account, API operation, Purchase Option
(e.g. Reserved), Service, and Tag.
Have visibility on your cost
AWS BUDGETS
Have visibility on your cost
AWS BUDGETS
Have visibility on your cost
AWS BUDGETS
Have visibility on your cost
AWS Cost Explorer
• Visualize your AWS cost
• Filters your AWS cost by AWS service, Linked Account,
Tag, month..
• Use same data as Detailed Billing Reports use
Have visibility on your cost
AWS Cost Explorer
• Log in to your AWS web console, click Cost Explorer
in Billing section
Have visibility on your cost
AWS Cost Explorer
Have visibility on your cost
AWS Cost Explorer – custom view
Have visibility on your cost
AWS Cost and Usage Report
• Cost and Usage Reports that provide you
with comprehensive data about products, pricing, and usage.
• Easily upload the reports to Amazon Redshift and
then run queries against the data using business
intelligence and data visualization tools including
Amazon QuickSight.
Have visibility on your cost
AWS Cost and Usage Report
AWS Cost and Usage Report
Have visibility on your cost
AWS Cost and Usage Report
Have visibility on your cost
AWS Cost and Usage Report & QuickSight !
Have visibility on your cost
AWS Cost and Usage Report & QuickSight !
Have visibility on your cost
AWS Cost and Usage Report & QuickSight !
Have visibility on your cost
Structure your Strategy
1. Have visibility on your cost
1. Budgets
2. Cost Explorer
3. Cost and Usage Report
2. Optimize AWS usage
1. Trusted Advisor
2. Right Sizing
3. Elasticity
3. Use Reserved Instance
Optimize AWS usage
AWS Trusted Advisor
• Trusted Advisor inspects your AWS environment and
finds opportunities to save money, improve system
performance and reliability, or help close security gaps.
• Since 2013, customers have viewed over 2.6 million best-
practice recommendations and realized over
$350 million in estimated cost reductions.
Optimize AWS usage
AWS Trusted Advisor – perspective of Cost Savings*
Checking Points
Amazon EC2 Reserved Instances Optimization
Low Utilization Amazon EC2 Instances
Idle Load Balancers
Underutilized Amazon EBS Volumes
Unassociated Elastic IP Addresses
Amazon RDS Idle DB Instances
* available vis Business and Enterprise Support
:: : No issue
: Investigation needed
: Immediate action needed
Optimize AWS usage
AWS Trusted Advisor – perspective of Cost Savings
Checking Points Remark
Amazon EC2 Reserved
Instances Optimization
calculates an optimal number of Partial Upfront Reserved Instances.
Low Utilization Amazon
EC2 Instances
alerts you if the daily CPU utilization was 10% or less and network I/O was 5 MB or less
on 4 or more days, out of total 14 days
Idle Load Balancers Checks your Elastic Load Balancing configuration for load balancers that are not
actively used
Underutilized Amazon
EBS Volumes
If a volume remains unattached or has very low write activity (excluding boot volumes)
for a period of time, the volume is probably not being used.
Unassociated Elastic IP
Addresses
Checks for Elastic IP addresses (EIPs) that are not associated with a running Amazon
Elastic Compute Cloud (Amazon EC2) instance.
Amazon RDS Idle DB
Instances
Checks the configuration of your Amazon Relational Database Service (Amazon RDS)
for any DB instances that appear to be idle.
Optimize AWS usage
Right Sizing
• Trusted Advisor determines if 10% or less CPU utilization
% and network I/O was 5 MB or less on 4 or more days in
14 days to identify underutilized EC2
• Metrics can be Low CPU, Low Bandwidth, Low I/O,
Days Alive, Estimated Cost, Name tag
AWS 프리미어 파트너 –
메가존 김성용 매니저를 소개합니다
클라우드 전문기업 메가존
1998년 설립, 19년간의 비즈니스 이력
본사 : 대한민국 서울
지사 : 베트남, 하노이
글로벌 파트너쉽 : + 20
320여명의 임직원
AWS와 함께하는 역사
2012 03 국내 최초 AWS 파트너십 체결
2015 04 국내 최초 및 세계 유일의 AWS 인증 서포트 파트너
2015 10 국내 최초 AWS 프리미어 파트너
2015 12 국내 최초 Marketing & Commerce Competency 파트너
2015 12 엔터프라이즈 서포트 계약 체결
국내 최초 엔터프라이즈 서포트 제공
주요 혜택 :
 수석 클라우드 지원 엔지니어
 어플리케이션 아키텍쳐 리뷰
 기술 지원 엔지니어(TAM) 일대일 지원
 고객 서비스 24/7/365 서포트
엔터프라이즈 서포트란?
“AWS 서포트 중 최상위 서포트 등급”
개발자 비즈니스 엔터프라이즈
보통
< 12시간
긴급
< 1시간 심각
< 15 분
서포트 등급별 최소응답시간
Gaming 고객사
비용절감을 위한 검증된 방법과 자료가 이미 많은데,
왜 어려울까요?
Chapter I.
1. 관리대상 과다
이미 관리 중인 IDC, AWS & 솔루션
 서버, 네트워크, 보안, 라이선스…
EC2 Instances
 1인 당 최소 1대 ~ 최대 1,000대
EBS
 1인 당 최소 1대 ~ 최대 600대
2. 관리방법 과다
 CPU Util%
 Memory usage
 EBS I/O
 Network I/O
 탄력성 (Elasticity)
 RI 활용률
 앞으로의 비즈니스 예측
3. 관리프로세스 확립의 어려움
 분석 리소스 확보 및 가시화
 이슈대상 정의 & 분석방법 정의
 이슈를 어떻게 해결할 것인가?
 해결책을 누가 어떻게 집행할 것인가?
 집행결과를 어떻게 보고할 것인가?
4. 이걸 다 조합하면..
 33 개 변수
 54 개 하위 변수
여기까지 검토했을 때의 표정
메가존과 AWS가 공개하는 AWS 비용 최적화 전략-메가존 김성용 매니저 및 AWS 이우상 매니저:: AWS Cloud Track 3 Gaming
Chapter II.
 CloudWatch와 빌링데이터를 활용한
이슈 탐지 및 해결책 도출하기
 할 수 있는 것 부터 하자!
Sample Case Study – “A 마케팅 회사”
“Right Sizing”
0
0.2
0.4
0.6
0.8
1
1.2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Right-Sizing
CPU : 100%
= 인스턴스 스펙다운
실제
CPU 사용률 :
다운그레이드
가능
0 ~ 25%
0
0.2
0.4
0.6
0.8
1
1.2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Right-Sizing
C3.xlarge
= 인스턴스 스펙다운
M3.large
Saving 36 %
=
C3.xlarge  m3.large
Right-Sizing : How to Convert
항목 ECU Ram. NetWork Price
C3.xlarge 14 7.5 Moderate 0.21
실제
사용량
3.5 0.5 Moderate 0.21
검색 조건:
ECU >= 3.5,
Ram >= 0.5,
Price < 0.21
= m3.largeM3.large 6.5 7.5 Moderate 0.133
36% 비용절감
Sample Case Study
“Right Sizing + Auto Scaling”
0
2
4
6
8
10
12
14
16
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Right Sizing + Auto Scaling
C3.xlarge, 24hrs
Auto-Scaling
+
Right Sizing
Without
Auto-Scaling
0
2
4
6
8
10
12
14
16
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Right Sizing + Auto Scaling
C3.xlarge, 24hrs
Saving 66%
M3.mid, 24hrs
M3.mid, 1hr
66% 비용 절감
“Right Sizing + Auto Scaling”
Sample Case Study
C3.xlarge
0시~20시
21시~22시
23시
사용시간
사용량
Optimization TO-BE
ECU
< 3.0
= 3.7
< 3.0
Right-Sizing
Right-Sizing
Auto Scaling
Right-Sizing
M3.mid x 1
M3.mid x 2
M3.mid x 1
3
3 + 3
3
5.04 1.68
Cost (Daily)ECUAS-IS
Sample Case Study
“Right Sizing + Auto Scaling + RI”
Right Sizing + Auto Scaling
C3.xlarge, 24hrs
Saving 66%
M3.mid, 24hrs
M3.mid, 1hr
0
2
4
6
8
10
12
14
16
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Right Sizing + Auto Scaling + RI
C3.xlarge, 24hrs
Saving 76%
M3.mid, 24hrs
+ RI
M3.mid, 1hr
0
2
4
6
8
10
12
14
16
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
76% 비용 절감
“Right Sizing + Auto Scaling + RI”
Sample Case Study
C3.xlarge
0시~20시
21시~22시
23시
사용시간
사용량
Optimization TO-BE
ECU
< 3.0
= 3.7
< 3.0
Right-Sizing(R)
Right-Sizing(R)
Auto Scaling
Right-Sizing(R)
M3.mid x 1
M3.mid x 1
M3.mid x 1
M3.mid x 1
3
3 + 3
3
5.04 1.2
Cost (Daily)ECUAS-IS
요약 들어갑니다~
C3.xlarge  M3 Cost Optimization Plans
Right Sizing
Auto-Scaling
RI
Saving
Plan 1 Plan 3 Plan 4 Plan 5
v v v
v v
v v v
30% 54% 66% 76%
Difficulty EASY Normal Hard Hard
Plan 2
v
36%
Normal
“이걸 언제 다 하지... 좀 더 간편하게
관리할 수 있으면 좋겠다.”
“Hyper Ops”
Demo - Right Sizing
Demo – RI 최적화
Reference : Right-Sizing
A 게임사
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
Before
After
49개 인스턴스 Right-Sizing.
$ 17,791.56, 34% 절감효과
Reference : RI Optimization
C3 M3 C4 M4 M1 C1 I2 D2 C4 t2
Ondemand 100%
AS-IS
RI 100%
 누적 성과와 미래 성과 분석
 계정, 인스턴스별로 ODI & RI 현황 파악
 RI를 통한 누적절감비용 보고 (월간 레포트)
 Ondemand -> RI 전환 시 추가 절감비용 시뮬레이션
 만료 예정 RI Alarm
B 게임사
Reference : RI management
 휴면 RI 색출
 휴면 원인 탐색
 휴면 상태를 활동상태로 전환
 RI 통합 & RI 분할 & AZ 변경 & VPC전환
 Saved Up To $20K
C 게임사
52,080
47,040
52,080
47,329
44,640
43,200
44,640
42,408 42,620 42,404
41,040
41,353
36,843
39,795
37,408 37,944
36,720
37,944
40,288
42,620 42,385
41,040
30,000
35,000
40,000
45,000
50,000
55,000
1 2 3 4 5 6 7 8 9 10 11
HOUR
MONTH
구매된 시간
사용된 시간
Use Reserved
Reserved Instance Pricing History
• AWS’s historical price drop
• History confirms that
on-demand price
does not go lower
than reserved instance price
* Source : cloudability.com
Use Reserved
* Source : cloudability.com
• Basics
• Structure your Strategy
• Conclusion
Use Reserved & Spot Instance
Overall Process
• Attach tag to all EC2s
• Monitor your usage / Alarm
• Find and remove underutilized resources
• Find right instance type
• Optimize using RI
• Define governance / Assign an owner

More Related Content

PDF
데이터 마이그레이션 AWS와 같이하기 - 김일호 솔루션즈 아키텍트:: AWS Cloud Track 3 Gaming
PDF
AWS를 통한 데이터 분석 및 처리의 새로운 혁신 기법 - 김윤건, AWS사업개발 담당:: AWS Summit Online Korea 2020
PDF
20150724 제10회 부산 모바일 포럼 - 클라우드컴퓨팅과 함께하는 아마존 웹 서비스
PDF
AWS를 활용한 금융권 hybrid cloud 구축하기 :: Felix Candelario :: AWS ...
PDF
AWS Media Day-AWS 기반의 미디어 & 엔터테인먼트 워크플로우 소개(Ben Masek 미디어 엔터테인먼트 부서 (M&E) 글로벌...
PDF
AWS Step Functions을 활용한 서버리스 앱 오케스트레이션
PDF
AWS 클라우드를 통한 교육 및 연구 혁신 - AWS Summit Seoul 2017
PDF
AWS Batch를 통한 손쉬운 일괄 처리 작업 관리하기 - 윤석찬 (AWS 테크에반젤리스트)
데이터 마이그레이션 AWS와 같이하기 - 김일호 솔루션즈 아키텍트:: AWS Cloud Track 3 Gaming
AWS를 통한 데이터 분석 및 처리의 새로운 혁신 기법 - 김윤건, AWS사업개발 담당:: AWS Summit Online Korea 2020
20150724 제10회 부산 모바일 포럼 - 클라우드컴퓨팅과 함께하는 아마존 웹 서비스
AWS를 활용한 금융권 hybrid cloud 구축하기 :: Felix Candelario :: AWS ...
AWS Media Day-AWS 기반의 미디어 & 엔터테인먼트 워크플로우 소개(Ben Masek 미디어 엔터테인먼트 부서 (M&E) 글로벌...
AWS Step Functions을 활용한 서버리스 앱 오케스트레이션
AWS 클라우드를 통한 교육 및 연구 혁신 - AWS Summit Seoul 2017
AWS Batch를 통한 손쉬운 일괄 처리 작업 관리하기 - 윤석찬 (AWS 테크에반젤리스트)

What's hot (14)

PPTX
인공지능 서비스를 활용한 여행 물류 산업의 고객 서비스 개인화 및 추천 기술 - 최원근 솔루션즈 아키텍트 :: AWS Travel and ...
PDF
AWS를 활용한 Big Data 실전 배치 사례 :: 이한주 :: AWS Summit Seoul 2016
PDF
Amazon Aurora (Debanjan Saha) - AWS DB Day
PDF
Premiers pas et bonnes pratiques sur Amazon AWS - Carlos Condé
PDF
AWS Summit Seoul 2015 - 일본 AWS 게임 고객사례 - Gungho, Grani, Nintendo를 중심으로
PDF
모두를 위한 MxNET - AWS Summit Seoul 2017
PDF
[판교에서 만나는 아마존웹서비스] AWS 클라우드 개요
PDF
2017 AWS DB Day | Amazon Database Migration Service (DMS) 소개 및 실습
PPTX
Understand AWS Pricing
PDF
AWS Lambda 활용의 모든 것! - AWS Summit Seoul 2017
PDF
일본 시골 개발자의 AWS 활용기 - AWS Summit Seoul 2017
PDF
Amazon EFS (Elastic File System) 이해하고사용하기
PDF
Construindo Aplicacoes Web e Mobile Escalaveis na AWS
PDF
교육의 진화, 클라우드는 어떤 역할을 하는가 :: Vincent Quah :: AWS Summit Seoul 2016
인공지능 서비스를 활용한 여행 물류 산업의 고객 서비스 개인화 및 추천 기술 - 최원근 솔루션즈 아키텍트 :: AWS Travel and ...
AWS를 활용한 Big Data 실전 배치 사례 :: 이한주 :: AWS Summit Seoul 2016
Amazon Aurora (Debanjan Saha) - AWS DB Day
Premiers pas et bonnes pratiques sur Amazon AWS - Carlos Condé
AWS Summit Seoul 2015 - 일본 AWS 게임 고객사례 - Gungho, Grani, Nintendo를 중심으로
모두를 위한 MxNET - AWS Summit Seoul 2017
[판교에서 만나는 아마존웹서비스] AWS 클라우드 개요
2017 AWS DB Day | Amazon Database Migration Service (DMS) 소개 및 실습
Understand AWS Pricing
AWS Lambda 활용의 모든 것! - AWS Summit Seoul 2017
일본 시골 개발자의 AWS 활용기 - AWS Summit Seoul 2017
Amazon EFS (Elastic File System) 이해하고사용하기
Construindo Aplicacoes Web e Mobile Escalaveis na AWS
교육의 진화, 클라우드는 어떤 역할을 하는가 :: Vincent Quah :: AWS Summit Seoul 2016
Ad

Similar to 메가존과 AWS가 공개하는 AWS 비용 최적화 전략-메가존 김성용 매니저 및 AWS 이우상 매니저:: AWS Cloud Track 3 Gaming (20)

PDF
[AWSome Day] Opening Keynote
PDF
AWS re:Invent re:Cap - 비용 최적화: 모범사례와 아키텍처 설계 기초편 - 이종남
PPTX
Cloud Expedition Technical1 - Día 1.pptx
PDF
SAP on AWS 이관사례로 알아보는 SAP 혁신 전략 - 이진욱, AWS SAP on AWS Solutions Architect
PDF
Amazon DocumentDB - Architecture 및 Best Practice (Level 200) - 발표자: 장동훈, Sr. ...
PPTX
KGC 2013 AWS session
PDF
AWS re:Invent re:Cap - 비용 최적화 - 모범사례와 아키텍처 설계 심화편 - 이원일
PPTX
Cloud cost optimization (AWS, GCP)
PPTX
Controle e otimização de custos na AWS
PDF
AWS Summit Berlin 2013 - Optimizing your AWS applications and usage to reduce...
PDF
12 Ways to Manage Cloud Costs and Optimize Cloud Spend
PDF
AWS Cloud cost optimization
PDF
Budget management with Cloud Economics | AWS Summit Tel Aviv 2019
PDF
Cost Optimisation on AWS
PDF
Best Practices and Resources to Effectively Manage and Optimize Your AWS Costs
PDF
Česko-Slovenský AWS Webinář 07 - Optimalizace nákladů v AWS
PDF
Six ways to reduce your AWS bill
PDF
ManageYourCostsAndGovernYourUsageOnAWS[1]
PDF
AWS Cost Optimization - JLM
PPTX
AWS EMEA Online Summit - Blending Spot and On-Demand instances to optimizing ...
[AWSome Day] Opening Keynote
AWS re:Invent re:Cap - 비용 최적화: 모범사례와 아키텍처 설계 기초편 - 이종남
Cloud Expedition Technical1 - Día 1.pptx
SAP on AWS 이관사례로 알아보는 SAP 혁신 전략 - 이진욱, AWS SAP on AWS Solutions Architect
Amazon DocumentDB - Architecture 및 Best Practice (Level 200) - 발표자: 장동훈, Sr. ...
KGC 2013 AWS session
AWS re:Invent re:Cap - 비용 최적화 - 모범사례와 아키텍처 설계 심화편 - 이원일
Cloud cost optimization (AWS, GCP)
Controle e otimização de custos na AWS
AWS Summit Berlin 2013 - Optimizing your AWS applications and usage to reduce...
12 Ways to Manage Cloud Costs and Optimize Cloud Spend
AWS Cloud cost optimization
Budget management with Cloud Economics | AWS Summit Tel Aviv 2019
Cost Optimisation on AWS
Best Practices and Resources to Effectively Manage and Optimize Your AWS Costs
Česko-Slovenský AWS Webinář 07 - Optimalizace nákladů v AWS
Six ways to reduce your AWS bill
ManageYourCostsAndGovernYourUsageOnAWS[1]
AWS Cost Optimization - JLM
AWS EMEA Online Summit - Blending Spot and On-Demand instances to optimizing ...
Ad

More from Amazon Web Services Korea (20)

PDF
[D3T1S01] Gen AI를 위한 Amazon Aurora 활용 사례 방법
PDF
[D3T1S06] Neptune Analytics with Vector Similarity Search
PDF
[D3T1S03] Amazon DynamoDB design puzzlers
PDF
[D3T1S04] Aurora PostgreSQL performance monitoring and troubleshooting by use...
PDF
[D3T1S07] AWS S3 - 클라우드 환경에서 데이터베이스 보호하기
PDF
[D3T1S05] Aurora 혼합 구성 아키텍처를 사용하여 예상치 못한 트래픽 급증 대응하기
PDF
[D3T1S02] Aurora Limitless Database Introduction
PDF
[D3T2S01] Amazon Aurora MySQL 메이저 버전 업그레이드 및 Amazon B/G Deployments 실습
PDF
[D3T2S03] Data&AI Roadshow 2024 - Amazon DocumentDB 실습
PDF
AWS Modern Infra with Storage Roadshow 2023 - Day 2
PDF
AWS Modern Infra with Storage Roadshow 2023 - Day 1
PDF
사례로 알아보는 Database Migration Service : 데이터베이스 및 데이터 이관, 통합, 분리, 분석의 도구 - 발표자: ...
PDF
Amazon Elasticache - Fully managed, Redis & Memcached Compatible Service (Lev...
PDF
Internal Architecture of Amazon Aurora (Level 400) - 발표자: 정달영, APAC RDS Speci...
PDF
[Keynote] 슬기로운 AWS 데이터베이스 선택하기 - 발표자: 강민석, Korea Database SA Manager, WWSO, A...
PDF
Demystify Streaming on AWS - 발표자: 이종혁, Sr Analytics Specialist, WWSO, AWS :::...
PDF
Amazon EMR - Enhancements on Cost/Performance, Serverless - 발표자: 김기영, Sr Anal...
PDF
Amazon OpenSearch - Use Cases, Security/Observability, Serverless and Enhance...
PDF
Enabling Agility with Data Governance - 발표자: 김성연, Analytics Specialist, WWSO,...
PDF
Amazon Redshift Deep Dive - Serverless, Streaming, ML, Auto Copy (New feature...
[D3T1S01] Gen AI를 위한 Amazon Aurora 활용 사례 방법
[D3T1S06] Neptune Analytics with Vector Similarity Search
[D3T1S03] Amazon DynamoDB design puzzlers
[D3T1S04] Aurora PostgreSQL performance monitoring and troubleshooting by use...
[D3T1S07] AWS S3 - 클라우드 환경에서 데이터베이스 보호하기
[D3T1S05] Aurora 혼합 구성 아키텍처를 사용하여 예상치 못한 트래픽 급증 대응하기
[D3T1S02] Aurora Limitless Database Introduction
[D3T2S01] Amazon Aurora MySQL 메이저 버전 업그레이드 및 Amazon B/G Deployments 실습
[D3T2S03] Data&AI Roadshow 2024 - Amazon DocumentDB 실습
AWS Modern Infra with Storage Roadshow 2023 - Day 2
AWS Modern Infra with Storage Roadshow 2023 - Day 1
사례로 알아보는 Database Migration Service : 데이터베이스 및 데이터 이관, 통합, 분리, 분석의 도구 - 발표자: ...
Amazon Elasticache - Fully managed, Redis & Memcached Compatible Service (Lev...
Internal Architecture of Amazon Aurora (Level 400) - 발표자: 정달영, APAC RDS Speci...
[Keynote] 슬기로운 AWS 데이터베이스 선택하기 - 발표자: 강민석, Korea Database SA Manager, WWSO, A...
Demystify Streaming on AWS - 발표자: 이종혁, Sr Analytics Specialist, WWSO, AWS :::...
Amazon EMR - Enhancements on Cost/Performance, Serverless - 발표자: 김기영, Sr Anal...
Amazon OpenSearch - Use Cases, Security/Observability, Serverless and Enhance...
Enabling Agility with Data Governance - 발표자: 김성연, Analytics Specialist, WWSO,...
Amazon Redshift Deep Dive - Serverless, Streaming, ML, Auto Copy (New feature...

Recently uploaded (20)

PDF
Machine learning based COVID-19 study performance prediction
PDF
Spectral efficient network and resource selection model in 5G networks
DOCX
The AUB Centre for AI in Media Proposal.docx
PPTX
Big Data Technologies - Introduction.pptx
PDF
Reach Out and Touch Someone: Haptics and Empathic Computing
PDF
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
PDF
The Rise and Fall of 3GPP – Time for a Sabbatical?
PDF
Review of recent advances in non-invasive hemoglobin estimation
PPTX
Detection-First SIEM: Rule Types, Dashboards, and Threat-Informed Strategy
PDF
Per capita expenditure prediction using model stacking based on satellite ima...
PDF
Agricultural_Statistics_at_a_Glance_2022_0.pdf
PPTX
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
PDF
Electronic commerce courselecture one. Pdf
PDF
NewMind AI Monthly Chronicles - July 2025
PDF
Empathic Computing: Creating Shared Understanding
PPTX
A Presentation on Artificial Intelligence
PDF
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
PDF
Approach and Philosophy of On baking technology
PDF
Unlocking AI with Model Context Protocol (MCP)
Machine learning based COVID-19 study performance prediction
Spectral efficient network and resource selection model in 5G networks
The AUB Centre for AI in Media Proposal.docx
Big Data Technologies - Introduction.pptx
Reach Out and Touch Someone: Haptics and Empathic Computing
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
The Rise and Fall of 3GPP – Time for a Sabbatical?
Review of recent advances in non-invasive hemoglobin estimation
Detection-First SIEM: Rule Types, Dashboards, and Threat-Informed Strategy
Per capita expenditure prediction using model stacking based on satellite ima...
Agricultural_Statistics_at_a_Glance_2022_0.pdf
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
Dropbox Q2 2025 Financial Results & Investor Presentation
Electronic commerce courselecture one. Pdf
NewMind AI Monthly Chronicles - July 2025
Empathic Computing: Creating Shared Understanding
A Presentation on Artificial Intelligence
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
Approach and Philosophy of On baking technology
Unlocking AI with Model Context Protocol (MCP)

메가존과 AWS가 공개하는 AWS 비용 최적화 전략-메가존 김성용 매니저 및 AWS 이우상 매니저:: AWS Cloud Track 3 Gaming

  • 1. 메가존과 AWS가 공개하는 AWS 비용 최적화 전략 김성용 매니져, MegaZone 이우상 매니져, AWS
  • 2. • Basics • Structure your Strategy • Conclusion
  • 3. Before Basics.. On and Off Fast Growth Predictable peaksVariable peaks WASTE (비용 낭비) CUSTOMER DISSATISFACTION (고객 이탈 발생)
  • 4. Before Basics.. Amazon EBS standa rd volume prices are lowered up to 50% for both storage and I/O requests. January 21, 2014 50% Amazon is reducing prices for Amazon ElasticCache cache nodes by an average of 34%. March 26, 2014 34% Amazon S3: We are reducing prices for Standard and Reduced Redundancy Storage, by an average of 51%. March 26, 2014 51% 1. Do Nothing 50 times since our inception in 2006. Recent price drops included…
  • 5. Before Basics.. 2. Do Almost Nothing CloudWatch, Trusted Advisor, Start Small
  • 6. Before Basics.. 3. Do Iterating Architecture Optimization
  • 7. Basics 4. 7 tips for Architecture Optimization • Turn off when not using (working 12 hours vs 24 hours) • Use Auto Scaling • Use Reserved Instance (1Y 35% ↓, 3Y 75% ↓) • Use Spot Instance (up to 90% ↓) • Use S3 options (S3, IA, Glacier → 3¢, 1.3 ¢, 0.7 ¢)* • Use AWS Dynamo DB (dynamic provisioning) • Remove un-needed component and utilize AWS managed services * Based on N. Virginia Region, for more details visit : https://aws.amazon.com/s3/pricing/?nc1=h_ls
  • 8. • Basics • Structure your Strategy • Conclusion
  • 9. Structure your Strategy 1. Have visibility on your cost 2. Optimize AWS usage 3. Use Reserved Instance
  • 10. Structure your Strategy 1. Have visibility on your cost 1. Budgets 2. Cost Explorer 3. Cost and Usage Report 2. Optimize AWS usage 1. Trusted Advisor 2. Right Sizing 3. Elasticity 3. Use Reserved Instance
  • 11. Have visibility on your cost AWS BUDGETS • Define and track budgets for your AWS costs, forecast your AWS costs • Budgets takes place on a fine-grained basis, with filtering or customization based on Availability Zone, Linked Account, API operation, Purchase Option (e.g. Reserved), Service, and Tag.
  • 12. Have visibility on your cost AWS BUDGETS
  • 13. Have visibility on your cost AWS BUDGETS
  • 14. Have visibility on your cost AWS BUDGETS
  • 15. Have visibility on your cost AWS Cost Explorer • Visualize your AWS cost • Filters your AWS cost by AWS service, Linked Account, Tag, month.. • Use same data as Detailed Billing Reports use
  • 16. Have visibility on your cost AWS Cost Explorer • Log in to your AWS web console, click Cost Explorer in Billing section
  • 17. Have visibility on your cost AWS Cost Explorer
  • 18. Have visibility on your cost AWS Cost Explorer – custom view
  • 19. Have visibility on your cost AWS Cost and Usage Report • Cost and Usage Reports that provide you with comprehensive data about products, pricing, and usage. • Easily upload the reports to Amazon Redshift and then run queries against the data using business intelligence and data visualization tools including Amazon QuickSight.
  • 20. Have visibility on your cost AWS Cost and Usage Report
  • 21. AWS Cost and Usage Report Have visibility on your cost
  • 22. AWS Cost and Usage Report Have visibility on your cost
  • 23. AWS Cost and Usage Report & QuickSight ! Have visibility on your cost
  • 24. AWS Cost and Usage Report & QuickSight ! Have visibility on your cost
  • 25. AWS Cost and Usage Report & QuickSight ! Have visibility on your cost
  • 26. Structure your Strategy 1. Have visibility on your cost 1. Budgets 2. Cost Explorer 3. Cost and Usage Report 2. Optimize AWS usage 1. Trusted Advisor 2. Right Sizing 3. Elasticity 3. Use Reserved Instance
  • 27. Optimize AWS usage AWS Trusted Advisor • Trusted Advisor inspects your AWS environment and finds opportunities to save money, improve system performance and reliability, or help close security gaps. • Since 2013, customers have viewed over 2.6 million best- practice recommendations and realized over $350 million in estimated cost reductions.
  • 28. Optimize AWS usage AWS Trusted Advisor – perspective of Cost Savings* Checking Points Amazon EC2 Reserved Instances Optimization Low Utilization Amazon EC2 Instances Idle Load Balancers Underutilized Amazon EBS Volumes Unassociated Elastic IP Addresses Amazon RDS Idle DB Instances * available vis Business and Enterprise Support :: : No issue : Investigation needed : Immediate action needed
  • 29. Optimize AWS usage AWS Trusted Advisor – perspective of Cost Savings Checking Points Remark Amazon EC2 Reserved Instances Optimization calculates an optimal number of Partial Upfront Reserved Instances. Low Utilization Amazon EC2 Instances alerts you if the daily CPU utilization was 10% or less and network I/O was 5 MB or less on 4 or more days, out of total 14 days Idle Load Balancers Checks your Elastic Load Balancing configuration for load balancers that are not actively used Underutilized Amazon EBS Volumes If a volume remains unattached or has very low write activity (excluding boot volumes) for a period of time, the volume is probably not being used. Unassociated Elastic IP Addresses Checks for Elastic IP addresses (EIPs) that are not associated with a running Amazon Elastic Compute Cloud (Amazon EC2) instance. Amazon RDS Idle DB Instances Checks the configuration of your Amazon Relational Database Service (Amazon RDS) for any DB instances that appear to be idle.
  • 30. Optimize AWS usage Right Sizing • Trusted Advisor determines if 10% or less CPU utilization % and network I/O was 5 MB or less on 4 or more days in 14 days to identify underutilized EC2 • Metrics can be Low CPU, Low Bandwidth, Low I/O, Days Alive, Estimated Cost, Name tag
  • 31. AWS 프리미어 파트너 – 메가존 김성용 매니저를 소개합니다
  • 32. 클라우드 전문기업 메가존 1998년 설립, 19년간의 비즈니스 이력 본사 : 대한민국 서울 지사 : 베트남, 하노이 글로벌 파트너쉽 : + 20 320여명의 임직원
  • 33. AWS와 함께하는 역사 2012 03 국내 최초 AWS 파트너십 체결 2015 04 국내 최초 및 세계 유일의 AWS 인증 서포트 파트너 2015 10 국내 최초 AWS 프리미어 파트너 2015 12 국내 최초 Marketing & Commerce Competency 파트너 2015 12 엔터프라이즈 서포트 계약 체결
  • 34. 국내 최초 엔터프라이즈 서포트 제공 주요 혜택 :  수석 클라우드 지원 엔지니어  어플리케이션 아키텍쳐 리뷰  기술 지원 엔지니어(TAM) 일대일 지원  고객 서비스 24/7/365 서포트 엔터프라이즈 서포트란? “AWS 서포트 중 최상위 서포트 등급” 개발자 비즈니스 엔터프라이즈 보통 < 12시간 긴급 < 1시간 심각 < 15 분 서포트 등급별 최소응답시간
  • 36. 비용절감을 위한 검증된 방법과 자료가 이미 많은데, 왜 어려울까요? Chapter I.
  • 37. 1. 관리대상 과다 이미 관리 중인 IDC, AWS & 솔루션  서버, 네트워크, 보안, 라이선스… EC2 Instances  1인 당 최소 1대 ~ 최대 1,000대 EBS  1인 당 최소 1대 ~ 최대 600대
  • 38. 2. 관리방법 과다  CPU Util%  Memory usage  EBS I/O  Network I/O  탄력성 (Elasticity)  RI 활용률  앞으로의 비즈니스 예측
  • 39. 3. 관리프로세스 확립의 어려움  분석 리소스 확보 및 가시화  이슈대상 정의 & 분석방법 정의  이슈를 어떻게 해결할 것인가?  해결책을 누가 어떻게 집행할 것인가?  집행결과를 어떻게 보고할 것인가?
  • 40. 4. 이걸 다 조합하면..  33 개 변수  54 개 하위 변수
  • 43. Chapter II.  CloudWatch와 빌링데이터를 활용한 이슈 탐지 및 해결책 도출하기  할 수 있는 것 부터 하자!
  • 44. Sample Case Study – “A 마케팅 회사” “Right Sizing”
  • 45. 0 0.2 0.4 0.6 0.8 1 1.2 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Right-Sizing CPU : 100% = 인스턴스 스펙다운 실제 CPU 사용률 : 다운그레이드 가능 0 ~ 25%
  • 46. 0 0.2 0.4 0.6 0.8 1 1.2 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Right-Sizing C3.xlarge = 인스턴스 스펙다운 M3.large Saving 36 %
  • 47. = C3.xlarge  m3.large Right-Sizing : How to Convert 항목 ECU Ram. NetWork Price C3.xlarge 14 7.5 Moderate 0.21 실제 사용량 3.5 0.5 Moderate 0.21 검색 조건: ECU >= 3.5, Ram >= 0.5, Price < 0.21 = m3.largeM3.large 6.5 7.5 Moderate 0.133 36% 비용절감
  • 48. Sample Case Study “Right Sizing + Auto Scaling”
  • 49. 0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Right Sizing + Auto Scaling C3.xlarge, 24hrs Auto-Scaling + Right Sizing Without Auto-Scaling
  • 50. 0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Right Sizing + Auto Scaling C3.xlarge, 24hrs Saving 66% M3.mid, 24hrs M3.mid, 1hr
  • 51. 66% 비용 절감 “Right Sizing + Auto Scaling” Sample Case Study C3.xlarge 0시~20시 21시~22시 23시 사용시간 사용량 Optimization TO-BE ECU < 3.0 = 3.7 < 3.0 Right-Sizing Right-Sizing Auto Scaling Right-Sizing M3.mid x 1 M3.mid x 2 M3.mid x 1 3 3 + 3 3 5.04 1.68 Cost (Daily)ECUAS-IS
  • 52. Sample Case Study “Right Sizing + Auto Scaling + RI”
  • 53. Right Sizing + Auto Scaling C3.xlarge, 24hrs Saving 66% M3.mid, 24hrs M3.mid, 1hr 0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
  • 54. Right Sizing + Auto Scaling + RI C3.xlarge, 24hrs Saving 76% M3.mid, 24hrs + RI M3.mid, 1hr 0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
  • 55. 76% 비용 절감 “Right Sizing + Auto Scaling + RI” Sample Case Study C3.xlarge 0시~20시 21시~22시 23시 사용시간 사용량 Optimization TO-BE ECU < 3.0 = 3.7 < 3.0 Right-Sizing(R) Right-Sizing(R) Auto Scaling Right-Sizing(R) M3.mid x 1 M3.mid x 1 M3.mid x 1 M3.mid x 1 3 3 + 3 3 5.04 1.2 Cost (Daily)ECUAS-IS
  • 57. C3.xlarge  M3 Cost Optimization Plans Right Sizing Auto-Scaling RI Saving Plan 1 Plan 3 Plan 4 Plan 5 v v v v v v v v 30% 54% 66% 76% Difficulty EASY Normal Hard Hard Plan 2 v 36% Normal
  • 58. “이걸 언제 다 하지... 좀 더 간편하게 관리할 수 있으면 좋겠다.” “Hyper Ops”
  • 59. Demo - Right Sizing
  • 60. Demo – RI 최적화
  • 61. Reference : Right-Sizing A 게임사 $0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 Before After 49개 인스턴스 Right-Sizing. $ 17,791.56, 34% 절감효과
  • 62. Reference : RI Optimization C3 M3 C4 M4 M1 C1 I2 D2 C4 t2 Ondemand 100% AS-IS RI 100%  누적 성과와 미래 성과 분석  계정, 인스턴스별로 ODI & RI 현황 파악  RI를 통한 누적절감비용 보고 (월간 레포트)  Ondemand -> RI 전환 시 추가 절감비용 시뮬레이션  만료 예정 RI Alarm B 게임사
  • 63. Reference : RI management  휴면 RI 색출  휴면 원인 탐색  휴면 상태를 활동상태로 전환  RI 통합 & RI 분할 & AZ 변경 & VPC전환  Saved Up To $20K C 게임사 52,080 47,040 52,080 47,329 44,640 43,200 44,640 42,408 42,620 42,404 41,040 41,353 36,843 39,795 37,408 37,944 36,720 37,944 40,288 42,620 42,385 41,040 30,000 35,000 40,000 45,000 50,000 55,000 1 2 3 4 5 6 7 8 9 10 11 HOUR MONTH 구매된 시간 사용된 시간
  • 64. Use Reserved Reserved Instance Pricing History • AWS’s historical price drop • History confirms that on-demand price does not go lower than reserved instance price * Source : cloudability.com
  • 65. Use Reserved * Source : cloudability.com
  • 66. • Basics • Structure your Strategy • Conclusion
  • 67. Use Reserved & Spot Instance Overall Process • Attach tag to all EC2s • Monitor your usage / Alarm • Find and remove underutilized resources • Find right instance type • Optimize using RI • Define governance / Assign an owner