Inquery#

After tutorial all participants are kindly asked to fill this Inquiry form(doc) or Inquiry form(pdf) .

Introduction and Life Demo of MAD, MaMe and EW with with canals simulation #

• PPT intro presentation and pdf version
• During the demo working with 2 segments connected with the gate will be shown

Assigments: Working with 3-segment canal simulation#

Note: We recommend to use Firefox browser

Step 1: Multiscale Application Designer - playing with MML

1.Open https://gs2.mapper-project.eu/mad

2.Push "Open from repository" button as indicated above

3.Choose "Canal 3 segments" MML and click Open

4. You should get gMML view of three SW1D models connected by gate and spillway

5. Now, click "Open in Experiment Workbench" button - new window with list of available executors (available resources) should pop up. This exercise will be conducted on grass1.man.poznan.pl machine

Step 2 - Executing experiment

1.Login to grass1.man.poznan.pl with appropriate login and password.

2.You should see generated experiment containing three snippets: 1)muscle, 2)canal visualizer and 3)mencoder movie maker. Click on the first snippet to see its contents

3.Modyfy parameters to customize canal application as follows:

#sets up the connection border side
cxa.env['SW1D_1B_001:borderconnection']='END'        
# Increase the number of iteration (computation)               
cxa.env['GLOBAL:iteration_number']=500    
#set up the rank of the second canal segment
cxa.env['SW1D_2B_005:segment_index']=2 
#the third canal segment will not produce a wave                
cxa.env['SW1D_1B_011:is_GAUSS']=false                     
# set up the rank of the third canal segment
cxa.env['SW1D_1B_011:segment_index']=4        
#set up the rank of the SpillWay           
 cxa.env['SpillWay_015:position_index']=3  
#set up the rank of the Gate                 
cxa.env['LinearGate1B_021:position_index']=1            
#set up water level of the third segment to see spillway effect
 cxa.env['SW1D_1B_011:water_level']=0.098                 

Note: if you draw MML from scratch ids (001, 005, 011) of the kernels can be different.

4. Now click run button of the first snippet as indicated in the picture below. The frame of the snippet should turn dark blue and the activity indicator should turn red.

5. Once simulation is finished the frame and the activity indicator should turn gray again. You should see output files appearing on the left side of the window.

6. Try to run next snippets in the same way. Second snippet should produce 500 images (one per iteration) in a Movie_Frame_Encoder_in_img_XXX directory

7. Third snippet produces movie from images and places it under the name Movie_Frame_Encoder_out_avi_movie_XXX.

8. Movie can be fetched using the browser and viewed by movie player installed in your system. In the movie you can observe how different components of the simulation (shallow water segments, gate and spillway) participate together in a water flow.

Step 3 Running distributed snippet execution

In this step we will show how to distribute snippet execution. The first snippet (MUSCLE) will be run on grass machine as in step 2. The second and the third snippet (CanalVisualiser and Mencoder movie maker) will be run on the second machine.

1. Log in to another machine from GridSpace Experiment workbench. Click on the small black arrow in the rigth top corner of the window as indicated in the picture below. Choose another executor.

2. Choose collage-eh.cyfronet.pl executor.

3. Now, you can switch executors assigned to snippets. Do it for CanalVisualiser and Mencoder snippets

3. You can also switch between executors and see file systems on each of them. Please do it as indicated on pictures below.

4. In order to transfer files between different executors, one must point out its exact location. For our experiment click on the input of the second (CanalVisualizer) snippet as indicated below.

5. Open asset information window and put its URL on the source executor (grass1) at the bottom of the window. You can click on "pencil" icon and choose Canal_Visualizer_in_data_027 directory from the list

You can also manually enter the url.

file://SSH-grass1.man.poznan.pl/mschool25/Canal_Visualizer_in_data_027/

The result should look like this:

6. Now run second and third snippet on collage-eh.cyfronet.pl executor. You should get output files on the left side of the window