Exercise 1: Ranking of Computing Services based on Service Efficiency metric

Option A

The participants can have access to the GUI (Cloud-Bench) of the Benchmarking Suite tool through the following link: https://docs.google.com/document/d/1DWicOAGjIpOdHO2zAz3K6AnrmnyDTW-1ITm8RCyMJHI/edit

 

Option B

  • Download Eclipse IDE (Luna 4.4.0 is recommended)

Perform Benchmarking GUI

Download the last version of the tool:

  • Download and install MySql Server.

Run the sql dump script for the creation of the benchmarking database: https://github.com/artist-project/ARTIST/blob/master/source/Tooling/pre-migration/Benchmarking%20Suite/MySqlDump.sql

Step 1: Import the WAR file

  • Import our binary file to Eclipse

In the main menu select File > Import > web > WAR; Browse to Cloud-bench.war file, choose your Target Environment andclick Finish.

The project will be associated to Java EE perspective.

 

Step 2: Deploy the WAR file

  • Application server to select

Right click on the just created web project > Run as > Run on Server; Select your Server, then click Finish.

 

 Step 3: Run the GUI

  • URL to access

Navigate via browser to http://server_ip_address:server_port/Cloud-Bench/reports

 

You have successfully performed the benchmark GUI interface


 

 Optional Activity (for advanced users)

If the participants are familiar with the benchmarking process they could download, install and execute on their own the benchmark tests on their pc. In this case the users should have an Amazon EC2 account,otherwise it is not possible for the users to complete the benchmarking process.

Instruction steps

  • Download the latest version of the tool:

https://github.com/artist-project/ARTIST/blob/master/binary/BenchmarkingSuite/Benchmarking%20Controller--3.0.0-0.tar.gz  (binary)

  • Download and install MySql Server.

Run the sql dump script for the creation of the benchmarking database: https://github.com/artist-project/ARTIST/blob/master/source/Tooling/pre-migration/Benchmarking%20Suite/MySqlDump.sql

Database credentials: user='root',passwd='test',db='cloudbench'

  • Use your Amazon EC2 account in order to create an AMI

 −        create a new instance from the EC2 dashboard(Amazon Linux AMI 2015.03 (PV) - ami-1ccae774)

−        save the new key (remember the related path and name; those information will be used in the configuration file of the benchmarking controller)

−        right click on the just created instance --> AMI --> Create image -->

  •   Download and install Eclipse IDE (Eclipse Luna 4.4.0 is recommended)

 

Step 1: Python Development Environment

  • Download PyDev
  • Open the PyDev Perspective
  • In the main menu select Window > Open Perspective > PyDev

 

Step 2: Working withPyDev

  • Create a PyDev project

You need a project to store our source code. In the main toolbar, click on the New PyDev Project button. Enter Controller  as project name, then click Finish.

 

Step 3: Import the binary of the tool

  • Import the Controller into created PyDev project

Right click on Controller project >  Import > Archive File. Click Next; Browse to Benchmarking Controller--3.0.0-0.tar.gz

 

Step 4: Benchmark MySql database connection

  • Database configuration

In the PyDev editor, access to src/eu/artist/benchmarking/parsers/database.py file and enter your benchmark database connection parameter. Default ones are the following:

MySQLdb.connect(host='127.0.0.1',user='root',passwd='test',db='cloudbench')

 

Step 5: Configure the Controller for a Cloud Provider

  • Example for Amazon EC2 (redHat instance, m1.large)

 

The next step is to create a new Cloud Provider configuration file. In the PyDev editor, access to cloud_providers folder and edit a new file amazon.conf. Similarly to example.conf file available in the folder, fulfill it with the following information:

cp_class = eu.artist.benchmarking.cloudproviders.amazon.AmazonProvider

access_id =  your Amazon EC2 access ID

secret_key =  your Amazon EC2 secret key

key_name =  your key name

key_path =  path to your local .pem key

 

[redhat-large]

platform=redhat

script_deploy_location = /home/ec2-user

output_file =/home/ec2-user/output.dat

image_id = your ami id

size_id = m1.large

vm_user = ec2-user

 

Step 6: Using the controller for the submission of ax existing workload on Amazon

  • Running a DaCapo workload on the Amazon instance

Right click  to test/tests.py > Run as > Python run

 

You have successfully performed the Benchmarking Controller; the performance data  related to the execution of a single workload (belonging to the DaCapo Suite tool) on m1.large Amazon EC2 instance, are now available in the benchmark database (dacapo table).