Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Self learning cloud controllers

1,152 views

Published on

Published in: Software
  • I pasted a website that might be helpful to you: ⇒ www.HelpWriting.net ⇐ Good luck!
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
  • You can ask here for a help. They helped me a lot an i`m highly satisfied with quality of work done. I can promise you 100% un-plagiarized text and good experts there. Use with pleasure! ⇒ www.HelpWriting.net ⇐
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here

Self learning cloud controllers

  1. 1. Self-Learning Cloud Controllers Pooyan Jamshidi, Amir Sharifloo, Claus Pahl, Andreas Metzger, Giovani Estrada IC4, Dublin City University, Ireland University of Duisburg-Essen, Germany Intel, Ireland pooyan.jamshidi@computing.dcu.ie Invited presentation at UFC, Brazil, Fortaleza
  2. 2. ~50% = wasted hardware Actual traffic Typical weekly traffic to Web-based applications (e.g., Amazon.com)
  3. 3. Problem 1: ~75% wasted capacity Actual demand Problem 2: customer lost Traffic in an unexpected burst in requests (e.g. end of year traffic to Amazon.com)
  4. 4. Really like this?? Auto-scaling enables you to realize this ideal on-demand provisioning Time   Demand   ?   Enacting change in the Cloud resources are not real-time
  5. 5. Capacity we can provision with Auto-Scaling A realistic figure of dynamic provisioning
  6. 6. 0 50 100 0 500 1000 1500 0 50 100 100 200 300 400 500 0 50 100 0 1000 2000 0 50 100 0 200 400 600
  7. 7. These quantitative values are required to be determined by the user ⇒  requires deep knowledge of application (CPU, memory, thresholds) ⇒  requires performance modeling expertise (when and how to scale) ⇒  A unified opinion of user(s) is required Amazon auto scaling Microsoft Azure Watch 9   Microsoft Azure Auto- scaling Application Block
  8. 8. Naeem Esfahani and Sam Malek, “Uncertainty in Self-Adaptive Software Systems” Pooyan Jamshidi, Aakash Ahmad, Claus Pahl, Muhammad Ali Babar , “Sources of Uncertainty in Dynamic Management of Elastic Systems”, Under Review
  9. 9. Uncertainty related to enactment latency: The same scaling action (adding/removing a VM with precisely the same size) took different time to be enacted on the cloud platform (here is Microsoft Azure) at different points and this difference were significant (up to couple of minutes). The enactment latency would be also different on different cloud platforms.
  10. 10. Ø Offline benchmarking Ø Trial-and-error Ø Expert knowledge Costly and not systematic A. Gandhi, P. Dube, A. Karve, A. Kochut, L. Zhang, Adaptive, “Model-driven Autoscaling for Cloud Applications”, ICAC’14 arrival  rate  (req/s)   95%  Resp.  :me  (ms)   400  ms     60  req/s  
  11. 11. RobusT2Scale Initial setting + elasticity rules + response-time SLA environment monitoring application monitoring scaling actions Fuzzy Reasoning Users Prediction/ Smoothing
  12. 12.      0 0.5 1 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Region  of   definite   satisfaction   Region  of   definite   dissatisfaction  Region  of   uncertain   satisfaction   Performance Index Possibility Performance Index Possibility words can mean different things to different people Different users often recommend different elasticity policies 0 0.5 1 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Type-2 MF Type-1 MF
  13. 13. Workload Response time 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 x2 uMembershipgrade 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 uMembershipgrade => => UMF LMF Embedded FOU mean sd
  14. 14. Rule   (𝒍)   Antecedents   Consequent   𝒄 𝒂𝒗𝒈 𝒍   Workload   Response-­‐ time   Normal   (-­‐2)   Effort   (-­‐1)   Medium   Effort   (0)   High   Effort   (+1)   Maximum   Effort  (+2)   1   Very  low   Instantaneous   7   2   1   0   0   -­‐1.6   2   Very  low   Fast   5   4   1   0   0   -­‐1.4   3   Very  low   Medium   0   2   6   2   0   0   4   Very  low   Slow   0   0   4   6   0   0.6   5   Very  low   Very  slow   0   0   0   6   4   1.4   6   Low   Instantaneous   5   3   2   0   0   -­‐1.3   7   Low   Fast   2   7   1   0   0   -­‐1.1   8   Low   Medium   0   1   5   3   1   0.4   9   Low   Slow   0   0   1   8   1   1   10   Low   Very  slow   0   0   0   4   6   1.6   11   Medium   Instantaneous   6   4   0   0   0   -­‐1.6   12   Medium   Fast   2   5   3   0   0   -­‐0.9   13   Medium   Medium   0   0   5   4   1   0.6   14   Medium   Slow   0   0   1   7   2   1.1   15   Medium   Very  slow   0   0   1   3   6   1.5   16   High   Instantaneous   8   2   0   0   0   -­‐1.8   17   High   Fast   4   6   0   0   0   -­‐1.4   18   High   Medium   0   1   5   3   1   0.4   19   High   Slow   0   0   1   7   2   1.1   20   High   Very  slow   0   0   0   6   4   1.4   21   Very  high   Instantaneous   9   1   0   0   0   -­‐1.9   22   Very  high   Fast   3   6   1   0   0   -­‐1.2   23   Very  high   Medium   0   1   4   4   1   0.5   24   Very  high   Slow   0   0   1   8   1   1   25   Very  high   Very  slow   0   0   0   4   6   1.6   Rule ()   Antecedents   Consequent   Work load   Respons e -time   -2   -1   0 +1   +2   12   Medium   Fast   2   5   3 0   0   -0.9   10 experts’ responses ​ 𝑅↑𝑙 : IF (the workload (​ 𝑥↓1 ) is ​​ 𝐹 ↓​ 𝑖↓1  , AND the response- time (​ 𝑥↓2 ) is ​​ 𝐺 ↓​ 𝑖↓2  ), THEN (add/remove ​ 𝑐↓𝑎𝑣𝑔↑𝑙  instances). ​ 𝑐↓𝑎𝑣𝑔↑𝑙 =​∑ 𝑢=1↑​ 𝑁↓𝑙 ▒​ 𝑤↓𝑢↑𝑙 × 𝐶 /∑𝑢=1↑​ 𝑁↓ Goal: pre-computations of costly calculations to make a runtime efficient elasticity reasoning based on fuzzy inference
  15. 15. Liang, Q., Mendel, J. M. (2000). Interval type-2 fuzzy logic systems: theory and design. Fuzzy Systems, IEEE Transactions on, 8(5), 535-550. Scaling Actions Monitoring Data
  16. 16. 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.5954 0.3797 𝑀 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.2212 0.0000 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 x2 u 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 u Monitoring data Workload Response time
  17. 17. 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.5954 0.3797 0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.9568 0.9377
  18. 18. Performance index ​ 𝑦↓𝑙 ,​ 𝑦↓𝑟 
  19. 19. 0 50 100 0 500 1000 1500 0 50 100 100 200 300 400 500 0 50 100 0 1000 2000 0 50 100 0 200 400 600 0 50 100 0 500 1000 0 50 100 0 500 1000
  20. 20. 0 10 20 30 40 50 60 70 80 90 100 -500 0 500 1000 1500 2000 Time (seconds) Numberofhits Original data betta=0.10, gamma=0.94, rmse=308.1565, rrse=0.79703 betta=0.27, gamma=0.94, rmse=209.7852, rrse=0.54504 betta=0.80, gamma=0.94, rmse=272.6285, rrse=0.70858 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Big spike Dual phase Large variations Quickly varying Slowly varying Steep tri phase 0 50 100 0 500 1000 1500 0 50 100 100 200 300 400 500 0 50 100 0 1000 2000 0 50 100 0 200 400 600 0 50 100 0 500 1000 0 50 100 0 500 1000 RootRelativeSquaredError
  21. 21. SUT Criteria Big spike Dual phase Large variations Quickly varying Slowly varying Steep tri phase RobusT2Scale 973ms 537ms 509ms 451ms 423ms 498ms 3.2 3.8 5.1 5.3 3.7 3.9 Overprovisioni ng 354ms 411ms 395ms 446ms 371ms 491ms 6 6 6 6 6 6 Under provisioning 1465ms 1832ms 1789ms 1594ms 1898ms 2194ms 2 2 2 2 2 2 SLA: ​ 𝒓 𝒕↓ 𝟗𝟓 ≤𝟔𝟎𝟎 𝒎𝒔 For every 10s control interval • RobusT2Scale is superior to under-provisioning in terms of guaranteeing the SLA and does not require excessive resources • RobusT2Scale is superior to over-provisioning in terms of guaranteeing required resources while guaranteeing the SLA
  22. 22. 0 0.02 0.04 0.06 0.08 0.1 alpha=0.1 alpha=0.5 alpha=0.9 alpha=1.0 RootMeanSquareError Noise level: 10%
  23. 23. Self-Learning Controller
  24. 24. Current Solution (my PhD Thesis + IC4 work on auto-scaling) Future Plan Updating K in MAPE-K @ RuntimeDesign-time Assistance Multi-cloud
  25. 25. Fuzzifier Inference Engine Defuzzifier Rule base Fuzzy Q-learning Cloud ApplicationMonitoring Actuator Cloud Platform Fuzzy Logic Controller Knowledge Learning AutonomicController 𝑟𝑡 𝑤𝑤,  𝑟𝑡,  𝑡ℎ,  𝑣𝑚 𝑠𝑎 system state system goal
  26. 26. RobusT2Scal e Learned rules FQL Monitoring Actuator Cloud Platform .fis L W W ElasticBench 𝑤,   𝑟𝑡 𝑤,   𝑟𝑡,     𝑡ℎ,   𝑣𝑚 𝑠𝑎 Load Generator C system state WCF REST 𝛾, 𝜂, 𝜀, 𝑟
  27. 27. Cloud Platform (PaaS)On-Premise P: Worker Role L: Web Role P: Worker Role P: Worker Role Cache M: Worker Role Results : Storag e Blackboard : Storage LG: Console Auto-scaling Logic (controller) Policy Enforcer 1112 10 LB: Load Balance r Queue Actuator Monitoring
  28. 28. 0 0.2 0.4 0.6 0.8 1 1.2 0 4 12 15 21 26 33 39 47 53 60 68 76 83 90 95 103 110 118 124 130 136 145 152 159 166 171 179 185 193 200 206 214 218 219 228 236 242 249 255 263 267 271 275 283 290 295 303 307 314 320 S1 S2 S3 S4 S5
  29. 29. 0 0.5 1 1.5 2 2.5 0 50 100 150 200 250 300 350 q(9,5) 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0 50 100 150 200 250 300 350 q(7,5) -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 0 50 100 150 200 250 300 350 q(9,3) -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0 50 100 150 200 250 300 350 q(3,3)
  30. 30. 0 1 2 3 4 5 6 7 8 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
  31. 31. 1 2 3 4
  32. 32. Challenge  1:  ~75%  wasted  capacity A c t u a l   d e m a n d Challenge  2:   customer  lost Fuzzifier Inference   Engine Defuzzifier Rule   base Fuzzy Q-­‐learning Cloud  ApplicationMonitoring Actuator Cloud  Platform Fuzzy  Logic   Controller Knowledge  Learning Autonomic  Controller 𝑟𝑡 𝑤 𝑤,𝑟𝑡,𝑡ℎ,𝑣𝑚 𝑠𝑎 system  state system  goal RobusT2Scale Learned  rules FQL Monitoring Actuator Cloud  Platform .fis L W W ElasticBench 𝑤, 𝑟𝑡 𝑤, 𝑟𝑡,    𝑡ℎ, 𝑣𝑚 𝑠𝑎 Load  Generator C system  state WCF REST 𝛾, 𝜂, 𝜀, 𝑟
  33. 33. http://computing.dcu.ie/~pjamshidi/PDF/SEAMS2014.pdf More Details? => http://www.slideshare.net/pooyanjamshidi/ Slides? => Thank you! Aakash AhmadClaus Pahl Soodeh Farokhi Amir Sharifloo Armin Balalaie https://github.com/pooyanjamshidi Code? => Hamed Jamshidi Nabor Mendonca Brian CarrollReza Teimourzadegan

×