WTF is Modeling, Anyway!?
?
0 likes?91 views
1. The document discusses different types of models and defines a performance model as a mathematical framework used to assess performance data by relating key metrics through equations. 2. It provides a simple example of a real performance model created to analyze why a homegrown application was taking over 60 seconds to launch. A logarithmic relationship between launch time and number of files was identified through data analysis and visual confirmation. 3. The one-line performance model developed - that launch time is equal to a proportionality constant times the log of the number of files - accurately modeled the data to within 5% and revealed the problem was too many unnecessary files in the file tree that could be pruned for a 300% performance improvement.
WTF is Modeling, Anyway!?
- 1. WTF is ¡°Modeling¡±, Anyway!? Video conversation with Boris Zibitsker on the BEZNext Channel Dr. Neil J. Gunther ¡ª @DrQz Performance Dynamics August 2, 2017 SM c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 1 / 14
- 2. c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 2 / 14
- 3. Types of Models This word ¡°model¡± is overloaded in both english and technology: UML software modeling model train set Kim Kardashian ?nancial/accounting models Amdahl¡¯s law statistical regression numerical mesh simulation benchmark workload simulation support vector machines convolutional neural nets We need to specify clearly and unambiguously which model c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 3 / 14
- 4. What is a Performance Model? A performance model is a mathematical framework used to assess the validity of performance data (an overlooked necessity) data + model = information 1 Select performance metrics as inputs: ¦Ë, R, S, Q, . . . 2 Model is a relationship between those metrics: Q = ¦Ë R 3 Model outputs are calculated metrics 4 Compare calculated metrics with (other) measured metrics 5 Repeat until satis?ed Can then project metric values into circumstances that are not measured or not measureable c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 4 / 14
- 5. A Real Simple Real Model c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 5 / 14
- 6. The Environment 5 Production Environment . . . S390 Robotic tape silos IBM AIX/SP-2 50 nodes IBM AIX/SP-2 50 nodes SP2 SP2 FDDI rings User Tek X-terminals c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 6 / 14
- 7. The Data Problem: Home-grown application could take > 60 seconds to launch IBM cluster would cost $millions to replace c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 7 / 14
- 8. Data Sample Table: Mean launch time in seconds Server Files Time Xfs1 8371 18.57 Xfs2 7113 16.72 NFS1 4781 17.01 NFS2 109 9.41 Observation: 109 ?les is nearly 128 = 27 Log2(128) = 7 is close to 9 seconds 4781 is near 4096 = 212 Log2(4096) = 12 is close to 17 seconds 8371 is near 8192 = 213 Log2(8192) = 13 is close to 18 seconds c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 8 / 14
- 9. Visual Con?rmation 1 10 100 1000 10000 5101520 Log?Linear Plot Log (Number of files) Meanlaunchtime(seconds) ? 2017 Performance Dynamics Data LSQ fit c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 9 / 14
- 10. Visual Con?rmation 0 2000 4000 6000 8000 10000 05101520 Log Model Number of files Meanlaunchtime(seconds) ? 2017 Performance Dynamics Data Model Data Model c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 10 / 14
- 11. One-Line Performance Model Mean launch time R: R = k log10(N) where N is the number of remote-server ?les and k = 4.57 is proportionality constant for base-10 logarithms Table: Log model of mean R times Remote server Measured seconds Log model %Error Xfs1 18.57 17.929948 3.446698 Xfs2 16.72 17.606686 -5.303144 NFS1 17.01 16.818079 1.128281 NFS2 9.41 9.312522 1.035894 Model is accurate to within 5% But where does logarithmic behavior come from? c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 11 / 14
- 12. To get a log you need a tree 27 To Get a Log, You Need a Tree ? ? ? ? ? ? 0 1 2 1 10 100 Level Number this is of this logarithm c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 12 / 14
- 13. Modeling Results 1 Saved $multi-million IBM SP2 cluster 2 SP2 replacement would NOT have solved anything 3 Problem caused by ¡°best practices¡± for system management 4 Performance management was completely overlooked 5 Font server held ¡«15000 ?les but only ¡«1000 needed 6 Simple log performance model told the whole story 7 Simple ?x with no CapEx cost: prune the tree! 8 300% performance win in shortened launch times! 9 Log model more about explanation than prediction/forecasting c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 13 / 14
- 14. Contact Information Performance Dynamics Company Castro Valley, California www.perfdynamics.com perfdynamics.blogspot.com facebook.com/PerformanceDynamics twitter.com/DrQz info@perfdynamics.com +1-510-537-5758 c 2018 Performance Dynamics WTF is ¡°Modeling¡±, Anyway!? August 2, 2017 14 / 14