Add a show option on path_requests_run.py

[EstherLerouzic]

When launching path_requests_run.py no detailed propagation is provided on the different computed paths. User needs to run separate transmission_main program to see the detail of the propagation, however the path might be different (constrained in path_requests_run and shortest path in transmission_main_example.py).

A 'show' tag on each line in the excel input service sheet could help to see the propagation details. The option might be reserved for excel input only.

Additionally enable constraints (include or exclude nodes) on transmission_main_example.py might be useful

[ggrammel]

agree. One use case I would like to get to is to run a k-shortest-path-routing between A and Z resulting into a list of path with their propagation details for further analysis. I am currently helping myself by using exclude constraints which is a little cumbersome but given the limited fiber density of optical networks still feasible. Adding the propagation details is really valuable.

[ojnas]

Just FYI, an implementation of k-shortest-paths and exclude nodes constraints have been proposed here:

https://review.gerrithub.io/c/Telecominfraproject/oopt-gnpy/+/494445 https://review.gerrithub.io/c/Telecominfraproject/oopt-gnpy/+/494458

@ggrammel, exactly the use case you described is supported at https://gnpy.app/.

[ggrammel]

Thanks @ojnas, this works quite nicely! One thing I observed when testing was that the Graph is not always corresponding to the simulation. (see screenshot attached). If it is some bug in in the library, Perhaps a "refresh" buttow would do:

[ojnas]

@ggrammel, that is strange. Do you mean that the graph and results are not updated when you run a new simulation? I can't reproduce this. Could you please describe the steps you have performed when this happens?

[ggrammel]

Re-checked and seems a Browser issue (see below)

Original sequence (firefox 77.0.1):

1. pick the CORONET_CONUS topology
2. select k=4
3. calculate path Atlanta Scranton ==> here I observed a strange routing around New-York area that might be correct, but looks a bit funny for the path-2/3 (if I remember well)
4. to analyze the routing I zoomed into the new-york area after selecting each path (k=1..4)
5. because zooming is a bit tricky (OK to zoom in, but couldn't figure out zooming out again in a controlled way), I decided to re run the scenario directly between New-York and Scranton
6. changed the source to New York
7. keeping k = 4
8. Running the simulation ... and here we are.

I could reproduce a similar but more subtle case right now. Running through the same steps as above. then

1. looking at k=1 between NY and Scranton. To do this I had to zoom the part of the map using the lens
2. switching to k=2. The figure didn't change, it still shows the shorter 200km link rather than the 354km one

Next test using Safar1 13.1.1:

• no problems, the thing is refreshing after 2-3s

Comparing the scenarios: when changing k=x, Safari scales always back to the original map. If the same is supposed to happen in the FF case, then the scaled-out figure shown below should disappear and be replaced by the global topology, but that doesn't happen.

[ojnas]

Ok, for me in Chrome it also scales back to the original map. I will try to investigate what is the problem with FF.

For manually zooming out to the original map you can either double click in the figure or click the autoscale button next to the +- zoom buttons.