iCubGenova09 (iRonCub3) S/N:000 – Communication via ethernet with the robot seems slow after fixing #1856

[gabrielenava]

Robot Name 🤖

iCubGenova09 (iRonCub3) S/N:000

Request/Failure description

After fixing issue https://github.com/robotology/icub-tech-support/issues/1856, we noted that the communication via ethernet with the robot seems, sometimes, "slow".

Detailed context

More specifically, we observed the following:

• when I ping the head, the communication seems fast as usual (~1 ms)
• however, yarpmanager is super slow. When we press "refresh", it takes various seconds to actually refresh the interface
• yarpmotorgui is very slow to open. it can takes even minutes to open. Sometimes, the gui is lagging
• online visualizer sometimes crashes, it is slow at receiving updated data, and often there is a huge lag.
• we are having difficulties communicating commands to the turbines and to the flight control, e.g., we have to press multiple times the "start" and "stop" buttons in order to start or stop the turbines and the flight control
• on the other hands, applications that runs directly on the head like the flight control are not lagging and are able to meet the required frequency.

based on all these observations, and on the fact that we started to notice these issues after fixing the ethernet port in https://github.com/robotology/icub-tech-support/issues/1856, I think that something happened since then. What I think is that most likely something else broken during the accident that broke the ethernet port, and it is affecting the robot behaviour since then. I would unmount the jetpack and check again the ethernet board if possible.

Additional context

cc @DanielePucci

How does it affect you?

Complexifies iRonCub experiments

[gabrielenava]

cc @nicktrem @HosameldinMohamed

[gabrielenava]

think problem seems solved after we discovered that yarp conf on icub-head was configured with the wifi address 10.0.0.2 instead of the ethernet address 10.0.2.2. After fixing this, we did not experienced again the delay I was mentioning on my comment. I think the issue can be closed for now, I'll open a new one in case the problem occurs again!