Ryu controller - simple_switch_13_stp does not recalculate the Tree after a failure?

Question

I am using mininet with a Ryu controller.

For the Ryu controller I am using the command:

ryu-manager simple_switch_stp_13.py

Concerning mininet I use the topology defined in the file Topo1.py:

from mininet.topo import Topo

class Project( Topo ):

   def __init__( self ):
       # Initialize topology
       Topo.__init__( self )

       # Add hosts
       h1 = self.addHost('h1')
       h2 = self.addHost('h2')

       # Add switches
       s1 = self.addSwitch('s1')
       s2 = self.addSwitch('s2')
       s3 = self.addSwitch('s3')
       s4 = self.addSwitch('s4')

       # Add links
       self.addLink(h1,s1)
       self.addLink(s1,s2)
       self.addLink(s2,s4)
       self.addLink(s3,s1)
       self.addLink(s3,s4)
       self.addLink(h2,s4)

topos = { 'myTopo': ( lambda: Project() )}

Then I run mininet with the command:

sudo mn --custom Topo1.py --topo=myTopo --switch ovsk --controller=remote --mac

The connection between them is perfect; I succeed to run a ping from host 1 to host 2. What I want to do is to cut the link with s1 and s2 (without stopping the ping between h1 and h2).

Secondly, the issues I have : According to https://osrg.github.io/ryu-book/en/Ryubook.pdf page 65 I used the command:

link s1 s2 down.

But, I can observe using wireshark that the controller sends Port_Mod, in reaction to the Port_Status, which asks to the switch to put the interface up again. The switch does it so the ping repass by the interface between s1 and s2. I'm not understanding why the controller sends this Port_Mod?

In https://osrg.github.io/ryu-book/en/Ryubook.pdf page 65 they do not have this issue.

Then, I tried the command:

py net.delLinkBetween(s1, s2)

The link is totally deleted, the controller is notified by a Port Status but the root, the cost to the root ... do not evolved at all. Then, the ping does not restart and does not pass by s3. I also do not get why ?

Do you know if the two observations I did are normal ? And if yes, do you know how to correct my experiment in order to have the expected behavior ?