Presentation is loading. Please wait.

Presentation is loading. Please wait.

AWS BEST PRACTICES Module 3: Costs Optimization July 2017.

Similar presentations


Presentation on theme: "AWS BEST PRACTICES Module 3: Costs Optimization July 2017."— Presentation transcript:

1 AWS BEST PRACTICES Module 3: Costs Optimization July 2017

2 Performance Optimization
AWS Best Practices Training Program Basic Approaches Course overview Basic infrastructure requirements Reliability and Performance overview Costs optimization overview Monitoring tools overview Useful tools overview Security in AWS Basic security principles and tools Access management Data encryption and protection Network security Costs Optimization The “Minimum Principle” and Auto Scaling Serverless architectures Container-based architectures Resources review and optimization Using Reserved Instances Using Spot Instances Performance Optimization AWS Service Limits Components selection and review Backup Components failure withstanding

3 Introduction

4 AWS COSTS ELEMENTS OVERVIEW
EC2 Instance (Compute) EBS Volumes EBS IOPS AWS VM Price Date Transfer Elastic IP Elastic LB Mind Other Services!

5 AWS EC2 Billing Modes On Demand: Pay-as-you-go approach.
Reserved Instances: Reserve capacities for 1 or 3 years. Actual usage rate does not matter No upfront – pay monthly during 1 year Partial upfront – pay a portion upfront, and the rest – during the reservation period All upfront – pay all upfront, only once, for the whole period Spot Instances: Pay 30-90% of On Demand, but be ready that the instance can be shut off:

6 AWS COSTS OPTIMIZATION: SHORT OVERVIEW
Project costs is one of the crucial challenges managers face. Uncontrolled and unreasonable usage of virtual resources on AWS can lead to unexpected expenses, that can be easily omitted by keeping in mind several simple rules: Use the “Minimum” principle: order the resources with the minimum capacity, necessary for your needs. You can always expand, when necessary. Select proper resources: the correct resource selection can provide you with the necessary results without extra expenses. Use auto scaling to dynamically adjust your infrastructure to your needs Use Serverless and Container based approaches. Utilize resources reasonably: monitor low utilized resources and adjust their usage. Stopped resources cost up to 70% less than running ones Use Reserved and Spot Instances: a Spot instance can cost up to 90% less than an on-demand one. Use schedules: automate the start/stop processes Set up Quotas: to track and control monthly expenses

7 Take Minimum … and Scale

8 HINT 1: Take the minimum… and scale
AWS allows you to order the minimum necessary resources, and scale them – either manually or automatically, according to your needs. T2: burstable performance instances M4/M3: general purpose instances C4/C3: compute optimized instances X1/R4/R3: memory optimized instances P2/G2/F1: accelerated computing instances I3: High I/O instances D2: Dense-storage instances

9 HINT 2: Use Serverless and Container-Based Apps
Amazon EC2 Container Service AWS Lambdas Automatic scaling Automatic provisioning No need to manage infrastructure No need to learn new language You can use any library, including native Bring your code $0.20 per million requests, 1M free Easily manage Docker containers Flexible container placement Designed for use with other AWS services Extensible Performance at scale Secure

10 Reasonable Resources Usage

11 HINT 3: Use resources reasonably
Turn Off the Lights Developer, test, training instances Use simple instance start and stop Or tear down and build up all together using AWS CloudFormation Instances are disposable! Cuts ONLY compute price!

12 Using Schedules Running VM Stopped VM
The Idea Running VM Stopped VM <show live schedules> Automate infrastructure management Minimize human factor Set up any convenient schedule Automatic check of schedules execution

13 HINT 3: Use resources reasonably
AWS Trusted Advisor aws.amazon.com/premiumsupport/trustedadvisor/ Free with Business or Enterprise Support (that EPAM has)

14 HINT 3: Use resources reasonably
Use Optimization Statistics on the Monitoring Page cloud.epam.com/maestro2/ui/monitoring

15 HINT 3: Use resources reasonably
Low Utilization report

16 Reserved and Spot Instances

17 HINT 4: Use Reserved Instances
Reserved instances can help cut costs for predictable infrastructures significantly:

18 HINT 5: Use Spot Instances
Spot Instances Rules Price based on supply/demand You choose your maximum price/hour Your instance is started if the Spot price is lower Your instance is terminated if the Spot price is higher, with 2 minutes notice But: You did plan for fault tolerance, didn’t you? Best Practice Use Bid Advisor Set the bid price closer to the price of the on- demand instance of the similar type Set up fault tolerance Spot block: guaranteed capacity for 1-6 hours with 30-50% discount

19 HINT 5: Use Spot Instances
Bid Advisor Use the Bid Advisor tool to check the profit of each type of spot instances, and to estimate the possibility of outbidding

20 HINT 5: Use Spot Instances
Use Cases Stateless web/app server fleets Testing Batch workload Web servers Continuous integration (CI) High performance computing (HPC) Grid computing Media rendering/transcoding See more at: aws.amazon.com/ec2/spot

21 Expenses Control Tools

22 EPAM Orchestrator Monthly Quotas: Overview
Monthly quotas specify the desired limit for Cloud Expenses within the month. In case several quota rules are specified, the narrowest (with smallest limit) is applied. ALL - the quota is applied cumulatively to all regions in which the project is activated. All AWS(EPAM/GCP/Azure) – the quota is applied cumulatively to all regions of an external provider, in which the project is activated. EACH - the quota is applied to each region in which the project is activated. EACH AWS(EPAM/GCP/Azure) – the quota is applied in each region of an external provider, in which the project is activated. Region by name – Specify a quota for a specific region

23 EPAM Orchestrator Quotas: Management
Review quota settings Update quota limits Specify quota depletion actions

24 EPAM Orchestration Resource Quotas
Item Description Limit Type Volumes Number of storage volumes created 5 Daily Volume size Size of each storage volume 500 GB Total volume size Total size of each volume 2.5 TB Instances Number of instances created 20 Resource utilization The cost of the utilized infrastructure 300 USD Monthly PM can set approvals for VM Creation!

25 Unusual Activities Report
EPAM Orchestrator analyzes your AWS activities and informs you in case unusual activity is detected. When a daily resource creation quota in AWS is exceeded on your project within an hour, Project Coordinators and Delivery managers get a notification via SMS and a respective with the details on these activities, and two action buttons: Approve: verify that the enormous resource creation activities are expected Reject: all the resources listed in the report will be terminated, and the related accounts deleted When limits are set, we can start controlling meeting them. This report reacts on limits exceeding If Approve or Reject button is not pressed by Project Coordinator or Delivery Manager within 15 minutes after the notification, EPAM Cloud Support team can take the responsibility to Reject resources creation and to take all the necessary hack-protection measures.

26 NEXT STEPS

27 Security Education: Online Course
Get a free course on AWS APN portal ( Pass an APN Course or a Certification Submit a Certificate Get a badge Get to the pro-list for further courses + Investigate our Web Site ( + Watch the Video Portal (

28 Issues Resolving Flow Please also feel free to address EPAM Cloud Consulting team

29 AWS Enterprise Support Access
24/7 customer support with less than 1 hour response time Response to critical events less than 15 minutes Support by Senior Cloud Support Engineers via , chat and phone in case of critical events Unlimited number of cases Dedicated technical Account Manager and Concierge Agent Free Infrastructure Event Management Service Access to AWS Trusted Advisor and Support API functions Regular communication including AWS resource usage reporting, monitoring, recommendations on infrastructure optimization and improvement Access to Amazon documentation

30 Home Work Get a free course on AWS APN portal ( Pass AWS TCO and Cloud Economics Course

31 Documentation Hybrid Cloud Guide Terms and Conditions
Cloud Security Policy The guide providing the details on integration with AWS and Azure. EPAM Cloud terms and conditions: terms definitions, parties responsibilities EPAM Cloud Security policies and approaches FAQ Cloud Glossary Cloud Consulting What’s New Release Notes Video Overview Address us if you have any questions! EPAM Cloud frequently asked questions See the whole documentation set on

32 Performance Optimization
AWS Best Practices Training Program Basic Approaches Course overview Basic infrastructure requirements Reliability and Performance overview Costs optimization overview Monitoring tools overview Useful tools overview Costs Optimization The “Minimum Principle” and Auto Scaling Serverless architectures Container-based architectures Resources review and optimization Using Reserved Instances Using Spot Instances NEXT: Security in AWS Basic security principles and tools Access management Data encryption and protection Network security Performance Optimization AWS Service Limits Components selection and review Backup Components failure withstanding

33 Thank you for attention!
Cloud in Yammer: EPAM Cloud Consulting team


Download ppt "AWS BEST PRACTICES Module 3: Costs Optimization July 2017."

Similar presentations


Ads by Google