cloud index tracking enabling predictable costs in cloud
play

Cloud Index Tracking: Enabling Predictable Costs in Cloud Spot - PowerPoint PPT Presentation

Cloud Index Tracking: Enabling Predictable Costs in Cloud Spot Markets Supreeth Shastri and David Irwin University of Massachusetts Amherst Spot Servers are gaining significance in the cloud Servers that may terminate anytime after an advance


  1. Cloud Index Tracking: Enabling Predictable Costs in Cloud Spot Markets Supreeth Shastri and David Irwin University of Massachusetts Amherst

  2. Spot Servers are gaining significance in the cloud Servers that may terminate anytime after an advance warning period

  3. Spot Servers are gaining significance in the cloud Servers that may terminate anytime after an advance warning period Cheap Expensive Cost Guaranteed, Not guaranteed, Not guaranteed, Non-revocable Non-revocable Revocable Availability

  4. Spot Servers are gaining significance in the cloud Servers that may terminate anytime after an advance warning period Cheap Expensive On-demand Reserved Cost Spot Guaranteed, Not guaranteed, Not guaranteed, Non-revocable Non-revocable Revocable Availability

  5. Spot Servers are gaining significance in the cloud Servers that may terminate anytime after an advance warning period Cheap Expensive On-demand Spot instances helped scale our Reserved clusters up by 4X during the discovery of the Higgs Boson Cost Spot Researchers built the largest HPC cluster in the cloud with Guaranteed, Not guaranteed, Not guaranteed, 1.1million vCPUs on EC2 spot Non-revocable Non-revocable Revocable Availability

  6. Spot server pricing bid level while low on average, it is characterized by variability and deliberate revocations

  7. Spot server pricing bid level while low on average, it is characterized by variability and deliberate revocations Predicting Spot Prices is an Active Area of Research Ability to compare servers, plan IT budgets, and avoid disruptive revocations

  8. Spot server pricing bid level while low on average, it is characterized by variability and deliberate revocations Predicting Spot Prices is an Active Area of Research Ability to compare servers, plan IT budgets, and avoid disruptive revocations 2015 2016 2017 2018 Bid [ SIGCOMM ] No-bid [ HotCloud ] Prob-Guarantee [ SC ] LSTM [ HPDC ] SpotOn [ SoCC ] Flint [ Eurosys ] Proteus [ EuroSys ] Tributary [ ATC ] Cumulon [ VLDB ] BOSS [ Infocom ] Exosphere [ SIGMETRICS ]

  9. Predicting Spot Prices is Important Prior work models individual spot server prices based on their historical spot price data

  10. Difficult Accurately Predicting Spot Prices is Important Prior work models individual spot server prices based on their historical spot price data

  11. Difficult Accurately Predicting Spot Prices is Important Prior work models individual spot server prices based on their historical spot price data 2 2-5 14 2 68 Regions Time Hardware Zones OS commitments config (datacenters) types (country, state)

  12. Difficult Accurately Predicting Spot Prices is Important Prior work models individual spot server prices based on their historical spot price data 7600 + = 2 2-5 14 2 68 Regions Time Hardware Zones OS worldwide markets commitments config (datacenters) types (country, state)

  13. Difficult Accurately Predicting Spot Prices is Important Prior work models individual spot server prices based on their historical spot price data 7600 + = 2 2-5 14 2 68 Regions Time Hardware Zones OS worldwide markets commitments config (datacenters) types (country, state) One size fits all No visibility into Limited correlation model is unlikely market internals with external variables

  14. Image credit: www.cnbc.com/mad-money/

  15. vs. Image credit: www.cnbc.com/mad-money/

  16. Key Insight: A Market-based Index for CLOUD vs. Image credit: www.cnbc.com/mad-money/

  17. Key Insight: A Market-based Index for CLOUD vs. Rather than focusing exclusively on predicting individual servers, cloud users should make decisions based on broader market indices Image credit: www.cnbc.com/mad-money/

  18. [ intuition for our hypothesis Cloud Index index construction methodology validation on Amazon EC2 [ techniques for predictability Index-tracking design of index-tracking by server hopping performance evaluation

  19. Underlying Characteristics of Large Cloud Platforms

  20. Underlying Characteristics of Large Cloud Platforms 1. Dependence of VMs Spot markets originating from the same physical machine family are not free from mutual interference

  21. Underlying Characteristics of Large Cloud Platforms 1. Dependence of VMs Spot markets originating from the same physical machine family are not free from mutual interference Not all spot markets could be individually modeled and predicted

  22. Underlying Characteristics of Large Cloud Platforms 1. Dependence of VMs 2. Stability of Idle Capacity Spot markets originating from the Aggregate idle VM capacity in public same physical machine family are not cloud datacenters tends to be stable free from mutual interference [SoCC 2014, SOSP 2017] Not all spot markets could be individually modeled and predicted

  23. Underlying Characteristics of Large Cloud Platforms 1. Dependence of VMs 2. Stability of Idle Capacity Spot markets originating from the Aggregate idle VM capacity in public same physical machine family are not cloud datacenters tends to be stable free from mutual interference [SoCC 2014, SOSP 2017] Not all spot markets could be If idle capacity were priced like commodity, its individually modeled and predicted clearing price will be stable and predictable

  24. Underlying Characteristics of Large Cloud Platforms 1. Dependence of VMs 2. Stability of Idle Capacity Spot markets originating from the Aggregate idle VM capacity in public same physical machine family are not cloud datacenters tends to be stable free from mutual interference [SoCC 2014, SOSP 2017] Not all spot markets could be If idle capacity were priced like commodity, its individually modeled and predicted clearing price will be stable and predictable We hypothesize that observing spot markets at aggregate levels (say, server family or datacenter levels) should lead to stable prices

  25. Constructing a Market Index for CLOUD

  26. Constructing a Market Index for CLOUD Characterizing an individual server i Price = P i , Memory = M i GB Compute = C i ECUs P i norm = P i √ ( C i . M i )

  27. Constructing a Market Index for CLOUD Characterizing an individual server i Characterizing a group of servers Price = P i , Memory = M i GB Average of normalized prices Compute = C i ECUs N P i norm Σ P i norm = P i i=1 Index-level = √ ( C i . M i ) N

  28. Constructing a Market Index for CLOUD Characterizing an individual server i Characterizing a group of servers Price = P i , Memory = M i GB Average of normalized prices Compute = C i ECUs N P i norm Σ P i norm = P i i=1 Index-level = √ ( C i . M i ) N Cloud index value represents the average price per unit of compute time for the selected group of servers

  29. Individual Server Level bid level

  30. Individual Server Level bid level Datacenter Level (US-West-1a)

  31. Individual Server Level Server Family Level (US-West-1a) bid level Datacenter Level (US-West-1a)

  32. Individual Server Level Server Family Level (US-West-1a) bid level Datacenter Level (US-West-1a) Price prediction is more accurate and stable at datacenter- and server family level than individual level

  33. [ intuition for our hypothesis Cloud Index index construction methodology validation on Amazon EC2 [ techniques for predictability Index-tracking design of index-tracking by server hopping performance evaluation

  34. Design elements

  35. Design elements Index-tracking in financial markets S&P 500 Vanguard ETFs Investments that match the returns of an index. Construct a portfolio such that its constituent items are same as those present in the index.

  36. Design elements Index-tracking in financial markets Server hopping in cloud markets S&P 500 Vanguard ETFs A container that automatically hops spot VMs Investments that match the returns of an index. as market conditions change [ SoCC 2017 ]. Construct a portfolio such that its constituent items are same as those present in the index. Increasing cost-efficiency, lowers revocations

  37. Index Tracking by Server Hopping Achieving index-level cost-efficiency despite market volatility

  38. Index Tracking by Server Hopping Achieving index-level cost-efficiency despite market volatility 1 Determine a broad set of candidate markets , and then compute its market index

  39. Index Tracking by Server Hopping Achieving index-level cost-efficiency despite market volatility 1 2 Host the application on Determine a broad set of candidate markets , and then a server that meets the compute its market index index-level cost-efficiency

  40. Index Tracking by Server Hopping Achieving index-level cost-efficiency despite market volatility 1 2 3 Host the application on If market conditions Determine a broad set of candidate markets , and then a server that meets the violate the index invariant, then compute its market index index-level cost-efficiency transparently hop to a better server

Recommend


More recommend