roborescue / adf-sample-agent-java

A sample team using RCRS Agent Development Framework
https://roborescue.github.io/rcrs-adf-sample/
BSD 3-Clause "New" or "Revised" License
11 stars 26 forks source link

Problems with mounting agents #14

Closed Ab4ndon closed 2 years ago

Ab4ndon commented 2 years ago

I want to consult a problem encountered in operation. After I executed the command bash launch.sh -all, the interface remained in loading, and the agent did not run successfully

image

gnardin commented 2 years ago

@charon-1-1 The console log shows that the agents have been connected. The % does not reach 100%, it will remain in 75% until the end of the simulation.

To run the simulation, you need to select the button Run in the Kernel GUI window.

Ab4ndon commented 2 years ago

I successfully ran the sample code following your instructions, but when I tried to switch the map, I found that the RUN button was missing from the Kernel GUI. Figure 1 shows the example code running with the Run button, and Figure 2 shows me trying to switch maps. I wonder if there is something wrong with my command to switch the map(The command I use is bash start-comprun.sh -m ../maps/kobe/map).Could you please give me some instructions for switching map? Figure1 Figure2

gnardin commented 2 years ago

@charon-1-1 To run the simulation you are using the wrong script. You should use start.sh in order to see the Run button because the start-comprun.sh uses the --autorun flag when running the kernel, which means that it launches the simulation after 3 or 5 minutes automatically. The start-comprun.sh is a script used to run the competitions in a remote machine.

The proper command to run the simulator with a different scenario, for example Kobe, is

cd scripts
./start.sh -m ../maps/kobe/map -c ../maps/kobe/config

Here I set both the map and the configuration files associated to the Kobe scenario. If you do not set the -m or the -c, then the default map and configurations are used.

Ab4ndon commented 2 years ago

Thanks for your help, the sample code is ready to run on the server. However, when we tried to run our team's original old code, we found that the agent failed to mount, but the old code worked well on the old server. Is this caused by the new server? How do we get old code to run successfully on a new server? In addition, does the new server and agent have the latest operation manual for reference? adferror

gnardin commented 2 years ago

There was a change on the Server Port number due to a conflict on MacOS. The simulator server does not use port 7000 anymore, but port 27931. This was the only change that should affect the execution of legacy code on the new server. All other changes were internal to the server infrastructure and should not impact the execution of legacy code.

Please let us know if you face any further problem.

Ab4ndon commented 2 years ago

I tried to modify the port, but the terminal was stuck and could not execute downward any more. Is there something wrong with my command? adferror2

gnardin commented 2 years ago

@charon-1-1 Since your legacy code is based on ADF, you need to use the new version of the ADF. You can replace the modules you implemented on the new ADF version.

I suggest, you download the new simulator server and ADF versions. You run the sample ADF as it is to see if it works. If so, you copy the legacy modules to the new ADF version.