SlideShare a Scribd company logo

Cost Optimization on AWS - Pop-up Loft Tel Aviv

•Download as PPTX, PDF•
•
Amazon Web Services
Amazon Web Services

This document discusses how to optimize costs when using AWS. It recommends: 1) Architecting for cost efficiency by "paying for what you think you need"; 2) Optimizing usage costs by "paying for what you use"; and 3) Taking advantage of benefits over time by "paying for what you really need". It provides examples of using the right instance types, reserved instances, spot instances, and services to reduce costs. It also recommends monitoring billing closely and using tools like Trusted Advisor and the TCO calculator to find additional savings.

Related slideshows

Trustpilot
TrustpilotTrustpilot
Trustpilot
 
Journey Through the AWS Cloud: Cost Optimisation
Journey Through the AWS Cloud: Cost OptimisationJourney Through the AWS Cloud: Cost Optimisation
Journey Through the AWS Cloud: Cost Optimisation
 
Understand AWS Pricing
Understand AWS PricingUnderstand AWS Pricing
Understand AWS Pricing
 
1 of 31
Cost Optimization on AWS Rolf Kersten Business Development Manager
A Couple Assumptions… 1. You’re using AWS… 2. You like it!! 3. But probably want to spend less
AWS Pricing Philosophy More AWS Usage More Infrastructure Economies of Scale Lower Infrastructure Costs Reduced Prices More Customers Ecosystem Global Footprint New Features New Services Infrastructure Innovation We pass the savings along to our customers in the form of low prices and continuous reductions 51
1. Architect for Cost Efficiency „Pay for what You think you need“ 2. Optimize Usage Costs „Pay for what you use“ 3. Take Advantage of Benefits over Time „Pay for what you really need“ Cost Optimization with AWS
1. Architect for Cost Efficiency 2. Optimize Usage Costs 3. Take Advantage of Benefits over Time
Proactively Monitor Your Account Billing Usage…
By Using Detailed Billing Reports and Enabling Billing Alerts…
…Alongside Cost Explorer http://amzn.to/1zHE2Fj
EC2 Instance Usage Reports Cost per Instance Type per Hour, Day or Month ..or… Instance Hours per Instance Type per Hour, Day or Month All Data can be exported More Information: http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/usage-reports.html
Fit your payment model to your business model: EC2 pricing plans On-Demand Instances Reserved Instances Spot Instances Pay as you go for computing power Flat hourly rate, no up-front commitments Pay an up-front fee for a capacity reservation and a lower hourly rate (up to 72% savings) 1-year or 3-year terms RI Marketplace: Buy RIs with short remaining terms Pay what you want for spare EC2 capacity: your instances run if your bid exceeds the Spot price Potential for large scale at low cost: When they’re available, take advantage of 1,000s of Spot Instances at up to 90% savings 10:00 10:05 10:10 10:15
Standard Reserved Instances • You commit to pay for an instance for one or three years • We reserve the capacity, and offer you considerable savings over On Demand during the term • You decide how you pay: • No Upfront (Just hourly charges for all hours of the term) • Partial Upfont (Upfront Payment, plus hourly charges for all hours of the term) • All Upfront (Upfront Payment only)
Self Service Reserved Instance Utilization Reports RI „Bucket“ On Demand - (Usage + amortized upfront fee) More Information: http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/usage-reports.html
Spot Instance Use Case: HPC Clusters On Demand • Increase efficiency of solar panels by analysing 205.000 materials • 156.314 computing cores used at AWS, all regions, for 18 hours • Bill: $33.000 http://blog.cyclecomputing.com/2013/11/back-to-the-future-121-petaflopsrpeak-156000-core-cyclecloud-hpc-runs-264-years-of-materials-science.html
• Base Price - On Demand: No Commitment, Pay only what you use, no capacity reservation but once started no interruption • 10% less - Scheduled Reserved Instances: Commit to Schedule of at least 1200 hours for one year, capacity reservation • 30-60% less - Standard Reserved Instances: Commit to one or three years, capacity reservation • 40-60% less - Spot Blocks: Bid for 2-6 hours blocks of uninterrupted time • Up to 85% less - Spot: Bid for instances, interrupted if market price higher than your bid price, 2mins advanced notice An EC2 Cost Optimization Option for every Need
Combine all the Options! 1. Start by using RIs for known/steady-state workloads 2. Set-up multiple autoscaling groups 3. Autoscale first using Spot first 4. If Spot instances are unavailable (or priced too high), use On-Demand 0 2 4 6 8 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 /Spot On Demand Spot Reserved Instance
Trusted Advisor: Enterprise Strength Monitoring and Optimization • Monitors and recommends optimizations for: – Cost – Security – Fault Tolerance – Performance • Available to customers with Business and Enterprise- level support • Now four checks available as free tier! http://aws.amazon.com/premiumsupport/trustedadvisor/ “Over the last couple years, we’ve sent more than 2.6 million Trusted Advisor notifications that have saved customers over US$350 million dollars.”
Trusted Advisor: Cost Optimization Tips
Leverage The Right Storage Classes… AWS Cloud Amazon Glacier Gateway Appliance/ AWS Storage Gateway Amazon S3 Block File On-premises Data Center Archive Backup Disaster Recovery Amazon EBS • Amazon S3 Reduced Redundancy ‒ 99.99% durability vs. 99.999999999% • Amazon S3 Infrequent Access ‒ For data accessed less than two times a month • Amazon Glacier ‒ Same durability as S3 ‒ 3 to 5 hours restore time ‒ Up to 65% savings ‒ Great for archiving, long-term backups and old data Amazon EFS (Preview)
…Alongside AWS Managed or Application Services… Elastic Load Balancing (ELB) Amazon Relational Database Service (RDS) Amazon ElastiCache Amazon DynamoDB Amazon Simple Queue Service (SQS) Amazon Simple Email Service (SES) Amazon Lambda Amazon Elastic MapReduce (EMR) Amazon Simple Notification Service (SNS) Amazon Kinesis Amazon Route 53 Amazon Redshift
To Offload Your Architecture… + Standard Setup • 4 x Medium Instances $193 • AWS Data Transfer 1 TB $92 • Total = $285 Optimized • 1 x Medium Instance $48 • CloudFront Data 1 TB $87 • CloudFront Requests (10M) $7.5 • Total = $142.5 50% 6X Cheaper Faster
Intro toAmazon EC2 Dedicated Infrastructure Comparing Dedicated Hosts to Standard EC2 Instances  Multi-tenant servers host instances for multiple customers’ workloads  AWS determines which hosts instances run on  You pay per instance  Single-tenant servers host instances for one customer’s workloads  You can determine which hosts instances run on  You pay per host Standard Amazon EC2 Instances Amazon EC2 Dedicated Hosts EC2 Instance Customer 1 Customer 3 Customer 5 Customer 2 Customer 4 Customer 6 Customer X Amazon EC2 Dedicated Hosts Comparing Dedicated Hosts to Standard EC2 Instances Dedicated Hosts give you license flexibility Amazon EC2 Dedicated Instances Dedicated Infrastructure payment options Dedicated Hosts On-Demand Pricing Dedicated Reserved Instances Pricing Leveraging the power of the cloud compliantly
Intro toAmazon EC2 Dedicated Infrastructure Dedicated Hosts give you license flexibility When running Microsoft Workloads, for example:  Instance-host affinity allows you to comply with the 90 day rule  Per-core and per-socket visibility enables reporting at the physical level  Placement controls ensure that instances run on the server you want them to, helping you meet stringent compliance and regulatory requirements  AWS Config allows you to record when instances start, stop, or are terminated for license reporting and true-ups Amazon EC2 Dedicated Hosts Comparing Dedicated Hosts to Standard EC2 Instances Dedicated Hosts give you license flexibility Amazon EC2 Dedicated Instances Dedicated Infrastructure payment options Dedicated Hosts On-Demand Pricing Dedicated Reserved Instances Pricing Leveraging the power of the cloud compliantly
1. Architect for Cost Efficiency 2. Optimize Usage Costs 3. Take Advantage of Benefits over Time
How do customers lower their TCO with AWS? 1 Source: IDC Whitepaper, sponsored by Amazon, “Quantifying the Business Value of Amazon Web Services” May 2015 “Average of 400 servers replaced per customer” Replace up-front capital expense with lower “pay for what you use” variable cost model 3 Periodic Price Reductions Economies of scale allow AWS to continually lower costs 4 Save more money as you grow bigger Tiered Pricing Volume Discounts 2 Pricing model choice to support variable & stable workloads On-Demand Reserved Spot Dedicated
Analysts have shown AWS reduces costs over long term Source: IDC, Quantifying the Business Value of Amazon Web Services (May, 2015)
…and that these benefits increase over time Source: IDC Business Value of AWS Accelerates over time According to IDC, this relationship between length of time using AWS and return is due to customers leveraging the more optimized environment to generate more applications along a learning curve. $1 Investment in AWS $8.40 in benefits At 60 Months of using AWS ~8X $3.50 in benefits $1 Investment in AWS At 36 Months of using AWS ~3X
Comparing TCO is not easy (But We’re Going to Try) ≠
Elements of Infrastructure TCO Diagram doesn’t include every cost item. E.g. software costs can include database, management, middle tier software costs. Facilities cost can include costs associated with upgrades, maintenance, building security, taxes etc. IT labor costs can include security admin and application admin costs. Hardware – Server, Rack Chassis PDUs, ToR Switches (+Maintenance) Hardware – Storage Disks, SAN/FC Switches Network Hardware – LAN Switches, Load Balancer Bandwidth costs Server Admin Virtualization Admin 1 2 3 4 Server Costs Storage Costs Network Costs IT Labor Costs Software - OS, Virtualization Licenses (+Maintenance) Storage Admin costs Network Admin Costs Facilities Cost Space Power Cooling Facilities Cost Space Power Cooling Facilities Cost Space Power Cooling
AWS TCO Calculator: Example Cost Savings http://awstcocalculator.com
1. Architect for Cost Efficiency „Pay for what You think you need“ 2. Optimize Usage Costs „Pay for what you use“ 3. Take Advantage of Benefits over Time „Pay for what you really need“ Summary: Cost Optimization with AWS
Rolf Kersten Head of EMEA Platform Business Development kerstenr@amazon.de

More Related Content

Cost Optimization on AWS - Pop-up Loft Tel Aviv

  • 1. Cost Optimization on AWS Rolf Kersten Business Development Manager
  • 2. A Couple Assumptions… 1. You’re using AWS… 2. You like it!! 3. But probably want to spend less
  • 3. AWS Pricing Philosophy More AWS Usage More Infrastructure Economies of Scale Lower Infrastructure Costs Reduced Prices More Customers Ecosystem Global Footprint New Features New Services Infrastructure Innovation We pass the savings along to our customers in the form of low prices and continuous reductions 51
  • 4. 1. Architect for Cost Efficiency „Pay for what You think you need“ 2. Optimize Usage Costs „Pay for what you use“ 3. Take Advantage of Benefits over Time „Pay for what you really need“ Cost Optimization with AWS
  • 5. 1. Architect for Cost Efficiency 2. Optimize Usage Costs 3. Take Advantage of Benefits over Time
  • 6. Proactively Monitor Your Account Billing Usage…
  • 7. By Using Detailed Billing Reports and Enabling Billing Alerts…
  • 9. EC2 Instance Usage Reports Cost per Instance Type per Hour, Day or Month ..or… Instance Hours per Instance Type per Hour, Day or Month All Data can be exported More Information: http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/usage-reports.html
  • 10. Fit your payment model to your business model: EC2 pricing plans On-Demand Instances Reserved Instances Spot Instances Pay as you go for computing power Flat hourly rate, no up-front commitments Pay an up-front fee for a capacity reservation and a lower hourly rate (up to 72% savings) 1-year or 3-year terms RI Marketplace: Buy RIs with short remaining terms Pay what you want for spare EC2 capacity: your instances run if your bid exceeds the Spot price Potential for large scale at low cost: When they’re available, take advantage of 1,000s of Spot Instances at up to 90% savings 10:00 10:05 10:10 10:15
  • 11. Standard Reserved Instances • You commit to pay for an instance for one or three years • We reserve the capacity, and offer you considerable savings over On Demand during the term • You decide how you pay: • No Upfront (Just hourly charges for all hours of the term) • Partial Upfont (Upfront Payment, plus hourly charges for all hours of the term) • All Upfront (Upfront Payment only)
  • 12. Self Service Reserved Instance Utilization Reports RI „Bucket“ On Demand - (Usage + amortized upfront fee) More Information: http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/usage-reports.html
  • 13. Spot Instance Use Case: HPC Clusters On Demand • Increase efficiency of solar panels by analysing 205.000 materials • 156.314 computing cores used at AWS, all regions, for 18 hours • Bill: $33.000 http://blog.cyclecomputing.com/2013/11/back-to-the-future-121-petaflopsrpeak-156000-core-cyclecloud-hpc-runs-264-years-of-materials-science.html
  • 14. • Base Price - On Demand: No Commitment, Pay only what you use, no capacity reservation but once started no interruption • 10% less - Scheduled Reserved Instances: Commit to Schedule of at least 1200 hours for one year, capacity reservation • 30-60% less - Standard Reserved Instances: Commit to one or three years, capacity reservation • 40-60% less - Spot Blocks: Bid for 2-6 hours blocks of uninterrupted time • Up to 85% less - Spot: Bid for instances, interrupted if market price higher than your bid price, 2mins advanced notice An EC2 Cost Optimization Option for every Need
  • 15. Combine all the Options! 1. Start by using RIs for known/steady-state workloads 2. Set-up multiple autoscaling groups 3. Autoscale first using Spot first 4. If Spot instances are unavailable (or priced too high), use On-Demand 0 2 4 6 8 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 /Spot On Demand Spot Reserved Instance
  • 16. Trusted Advisor: Enterprise Strength Monitoring and Optimization • Monitors and recommends optimizations for: – Cost – Security – Fault Tolerance – Performance • Available to customers with Business and Enterprise- level support • Now four checks available as free tier! http://aws.amazon.com/premiumsupport/trustedadvisor/ “Over the last couple years, we’ve sent more than 2.6 million Trusted Advisor notifications that have saved customers over US$350 million dollars.”
  • 17. Trusted Advisor: Cost Optimization Tips
  • 18. Leverage The Right Storage Classes… AWS Cloud Amazon Glacier Gateway Appliance/ AWS Storage Gateway Amazon S3 Block File On-premises Data Center Archive Backup Disaster Recovery Amazon EBS • Amazon S3 Reduced Redundancy ‒ 99.99% durability vs. 99.999999999% • Amazon S3 Infrequent Access ‒ For data accessed less than two times a month • Amazon Glacier ‒ Same durability as S3 ‒ 3 to 5 hours restore time ‒ Up to 65% savings ‒ Great for archiving, long-term backups and old data Amazon EFS (Preview)
  • 19. …Alongside AWS Managed or Application Services… Elastic Load Balancing (ELB) Amazon Relational Database Service (RDS) Amazon ElastiCache Amazon DynamoDB Amazon Simple Queue Service (SQS) Amazon Simple Email Service (SES) Amazon Lambda Amazon Elastic MapReduce (EMR) Amazon Simple Notification Service (SNS) Amazon Kinesis Amazon Route 53 Amazon Redshift
  • 20. To Offload Your Architecture… + Standard Setup • 4 x Medium Instances $193 • AWS Data Transfer 1 TB $92 • Total = $285 Optimized • 1 x Medium Instance $48 • CloudFront Data 1 TB $87 • CloudFront Requests (10M) $7.5 • Total = $142.5 50% 6X Cheaper Faster
  • 21. Intro toAmazon EC2 Dedicated Infrastructure Comparing Dedicated Hosts to Standard EC2 Instances  Multi-tenant servers host instances for multiple customers’ workloads  AWS determines which hosts instances run on  You pay per instance  Single-tenant servers host instances for one customer’s workloads  You can determine which hosts instances run on  You pay per host Standard Amazon EC2 Instances Amazon EC2 Dedicated Hosts EC2 Instance Customer 1 Customer 3 Customer 5 Customer 2 Customer 4 Customer 6 Customer X Amazon EC2 Dedicated Hosts Comparing Dedicated Hosts to Standard EC2 Instances Dedicated Hosts give you license flexibility Amazon EC2 Dedicated Instances Dedicated Infrastructure payment options Dedicated Hosts On-Demand Pricing Dedicated Reserved Instances Pricing Leveraging the power of the cloud compliantly
  • 22. Intro toAmazon EC2 Dedicated Infrastructure Dedicated Hosts give you license flexibility When running Microsoft Workloads, for example:  Instance-host affinity allows you to comply with the 90 day rule  Per-core and per-socket visibility enables reporting at the physical level  Placement controls ensure that instances run on the server you want them to, helping you meet stringent compliance and regulatory requirements  AWS Config allows you to record when instances start, stop, or are terminated for license reporting and true-ups Amazon EC2 Dedicated Hosts Comparing Dedicated Hosts to Standard EC2 Instances Dedicated Hosts give you license flexibility Amazon EC2 Dedicated Instances Dedicated Infrastructure payment options Dedicated Hosts On-Demand Pricing Dedicated Reserved Instances Pricing Leveraging the power of the cloud compliantly
  • 23. 1. Architect for Cost Efficiency 2. Optimize Usage Costs 3. Take Advantage of Benefits over Time
  • 24. How do customers lower their TCO with AWS? 1 Source: IDC Whitepaper, sponsored by Amazon, “Quantifying the Business Value of Amazon Web Services” May 2015 “Average of 400 servers replaced per customer” Replace up-front capital expense with lower “pay for what you use” variable cost model 3 Periodic Price Reductions Economies of scale allow AWS to continually lower costs 4 Save more money as you grow bigger Tiered Pricing Volume Discounts 2 Pricing model choice to support variable & stable workloads On-Demand Reserved Spot Dedicated
  • 25. Analysts have shown AWS reduces costs over long term Source: IDC, Quantifying the Business Value of Amazon Web Services (May, 2015)
  • 26. …and that these benefits increase over time Source: IDC Business Value of AWS Accelerates over time According to IDC, this relationship between length of time using AWS and return is due to customers leveraging the more optimized environment to generate more applications along a learning curve. $1 Investment in AWS $8.40 in benefits At 60 Months of using AWS ~8X $3.50 in benefits $1 Investment in AWS At 36 Months of using AWS ~3X
  • 27. Comparing TCO is not easy (But We’re Going to Try) ≠
  • 28. Elements of Infrastructure TCO Diagram doesn’t include every cost item. E.g. software costs can include database, management, middle tier software costs. Facilities cost can include costs associated with upgrades, maintenance, building security, taxes etc. IT labor costs can include security admin and application admin costs. Hardware – Server, Rack Chassis PDUs, ToR Switches (+Maintenance) Hardware – Storage Disks, SAN/FC Switches Network Hardware – LAN Switches, Load Balancer Bandwidth costs Server Admin Virtualization Admin 1 2 3 4 Server Costs Storage Costs Network Costs IT Labor Costs Software - OS, Virtualization Licenses (+Maintenance) Storage Admin costs Network Admin Costs Facilities Cost Space Power Cooling Facilities Cost Space Power Cooling Facilities Cost Space Power Cooling
  • 29. AWS TCO Calculator: Example Cost Savings http://awstcocalculator.com
  • 30. 1. Architect for Cost Efficiency „Pay for what You think you need“ 2. Optimize Usage Costs „Pay for what you use“ 3. Take Advantage of Benefits over Time „Pay for what you really need“ Summary: Cost Optimization with AWS
  • 31. Rolf Kersten Head of EMEA Platform Business Development kerstenr@amazon.de

Editor's Notes

  1. Amazon CloudWatch can help you monitor the charges on sending you email alerts when charges exceed a threshold you define
  2. Power, cooling, and facilities are often high level allocations by corporate facilities, making it difficult to tie to specific hardware Complete asset inventories and hardware utilization is unknown Limited mapping of applications to specific physical or virtual hardware Limited tracking/visibility into hardware utilization and performance (servers, racks, UPS, power.. Chargeback and allocation to business units are often not understood or actionable