Commit 0c62933e authored by Jean-Didier Totow's avatar Jean-Didier Totow
Browse files

performance model, persistent storage

parent 766ef22d
# Performance Model
A task under the [Morphemic project](https://www.morphemic.cloud/).
This is an experimental repository in which we try to predict the performance of a service. Our model expects as input some configuration variables and provides as output an estimated performance. Our model is trained on history data that contains the configuration variables for a service and the current real resource allocation in course of time.
![](images/Morphemic-Page-2.png)
## Directory *data/*
It's a tidy place where we maintain our datasets, as simple & clear as possible, in order to be clear for all what to use for their experiments. Descriptive titles will be extremely helpful.
### For the secure document dataset ###
Metrics: time,AvgResponseTime,nrOfInstances,AvgMemory,AvgCPU,numebrOfRequest,Latency
Variables: nrOfInstances (the name of the file contains the cpu variable and memory variable)
those values should added in the dataset as column.
This dataset has a very low quality, it is a good opportunity to apply some data imputation technics
time,served_request,request_rate,response_time,performance,cpu_usage,memory
1603489027.766,2176,422,653.5170038833338,0.6640990171963114,32.2,65511424
1603489037.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489047.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489057.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489067.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489077.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489087.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489097.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489107.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489117.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489127.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489137.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489147.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489157.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489167.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489177.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489187.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489197.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489207.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489217.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489227.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489237.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489247.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489257.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489267.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489277.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489287.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489297.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489307.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489317.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
1603489327.766,2381,489,591.143068772612,0.8069112625990748,38.9,66756608
Secure Documents application​
Experiment with Secure Documents application was run on the instances with 2 cores and 8 GB of memory. The number of users has generated automatically in the range of 0 to 10. A new number of users was chosen in the random interval of time according to Poisson distribution with lambda 3000 milliseconds and constant delay offset 30000. Each user generates one request. Reconfiguration was allowed, min number of machines was 1, maximum 10. The application was running for 2 hours.​
The file contains rows with the following data:​
Timestamp​ - the time when one request was sent by one user. Present in format YYYY-MM-DD HH:mm:ss
numebrOfRequest – number of requests which are sending requests to the application.​
AvgResponseTime – average response time in the last 30 sekunds.
nrOfInstances – number of components running in parallel in a given moment.
AvgMemory – average memory utilisation in the last 30 seconds​.
AvgCPU - average CPU utilisation in the last 30 seconds.
Secure Documents application​
Experiment with Secure Documents application was run on the instances with 4 cores and 16 GB of memory. The number of users has generated automatically in the range of 0 to 10. A new number of users was chosen in the random interval of time according to Poisson distribution with lambda 3000 milliseconds and constant delay offset 30000. Each user generates one request. Reconfiguration was allowed, min number of machines was 1, maximum 10. The application was running for 2 hours.​
The file contains rows with the following data:​
Timestamp​ - the time when one request was sent by one user. Present in format YYYY-MM-DD HH:mm:ss
numebrOfRequest – number of requests which are sending requests to the application.​
AvgResponseTime – average response time in the last 30 seconds​.
nrOfInstances – number of components running in parallel in a given moment.
AvgMemory – average memory utilisation in the last 30 seconds​.
AvgCPU - average CPU utilisation in the last 30 seconds​.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment