Closed osrf-migration closed 5 years ago
Original comment by Jon Fink (Bitbucket: jonfink-arl).
Some additional (related?) information is that even when ignition comes down correctly, it seems to leave the two roslaunch
commands that it begins running.
Original comment by Nate Koenig (Bitbucket: Nathan Koenig).
@chapulina , I think the stuck roslaunch
commands are the ros1 bridge. Do you happen to have any ideas there?
Original comment by Nate Koenig (Bitbucket: Nathan Koenig).
Original comment by Louise Poubel (Bitbucket: chapulina, GitHub: chapulina).
addition of a 3D Grid to the main window
Is that from the menu on the top-right? That has never worked, we should remove it from the menu.
leave the two roslaunch commands that it begins running
Do you know exactly which ones? I see 7 roslaunch
calls inside virtual_stix.ign
Original comment by Louise Poubel (Bitbucket: chapulina, GitHub: chapulina).
even when ignition comes down correctly, it seems to leave the two roslaunch commands that it begins running
Also, what do you mean by "comes down correctly"? When I Ctrl + C
the terminal, all processed are finished correctly. If I close the window, other processes stay alive though, but I wouldn't expect them to be brought down.
Original comment by Jon Fink (Bitbucket: jonfink-arl).
This is only happening when some part of ignition crashes, e.g., when adding the 3D grid via the menu in the top-right. I have experienced other times where the ign launch ...
command never finishes bringing up the simulator (I get the GUI window popping up but do not see the 3D scene). Ctrl+C at the terminal does not work properly at this point resulting in my issues above.
In short, it seems there may be something blocking the correct shutdown of executables that have been launched via ignition if one part of the ignition system crashes.
Original comment by Jon Fink (Bitbucket: jonfink-arl).
I’m still try to figure out a reliable way to reproduce - definitely intermittent.
Original comment by baila (Bitbucket: baila).
Hey Jon Fink (jonfink-arl) , I guess I am also running into something similar. I have added my comment with logs and screenshot at [https://osrf-migration.github.io/subt-gh-pages/#!/osrf/subt/issues/82/problems-with-examplesetup-tutorial (#82)](https://osrf-migration.github.io/subt-gh-pages/#!/osrf/subt/issues/82/problems-with-examplesetup-tutorial (#82))
Difference is that, I used competition.launch file in y experiment.
Please share if you have found any workaround.
Original comment by baila (Bitbucket: baila).
Hi Nate Koenig (Nathan Koenig) , I am adding one file with all the logs during the run of the cmd :
ign launch virtual_stix.ign robotName1:=X1 robotConfig1:=X1_SENSOR_CONFIG_1
I get the window with the map only but no robot spawned.
The issue should be fixed very soon…I am stuck completely. Can’t work after moving to ignition framework. :disappointed:
Original comment by Nate Koenig (Bitbucket: Nathan Koenig).
I think you need to update your ros1_ign_bridge
code. Are you building ros1_ign_bridge
from source or using debians?
Original comment by Alfredo Bencomo (Bitbucket: bencomo).
Reopen if it still crashes with latest ign packages and tunnel_circuit branch.
Original report (archived issue) by Jon Fink (Bitbucket: jonfink-arl).
The original report had attachments: error_log.txt
Running into issues where Ignition crashes with a segfault and hangs (requiring many processes to be brought down by hand). Happens under a variety of circumstances, in this case with the addition of a 3D Grid to the main window (after
ign launch -v 4 virtual_stix.ign robotName1:=X1 robotConfig1:=X1_SENSOR_CONFIG_4
hangs during start, will create another issue for that).Relevant output includes:
At this point the command hangs at the terminal and does not respond to Ctrl+C. Killing the
ign
process with-9
leaves a number of processes still running that need to be brought down by hand.