search

USC-ACTLab / crazyswarm

A Large Quadcopter Swarm
MIT License
320 stars 315 forks source link

[crazyswarm_server-1] process has died when using the hover_swarm.launch #147

Closed lkyfly0765 closed 5 years ago

lkyfly0765 commented 5 years ago

We use the optitrack as mocap, and we have configured the Motive correctly. We follow the instructions on https://crazyswarm.readthedocs.io/en/latest/usage.html . The related yaml files has been changed correctly and the launch file has been configured as :

firmware: "bitcraze" # one of "crazyswarm", "bitcraze" broadcast_address: "E7E7E7E7E7" world_frame: "/world" genericLogTopics: ["log1"] genericLogTopicFrequencies: [10]

genericLogTopic_log1_Variables: ["ekfprof.usec_setup", "ekfprof.usec_innov", "ekfprof.usec_gain", "ekfprof.usec_corr", "ekfprof.usec_cov"]

  # genericLogTopic_log1_Variables: ["profiling.usec_ekf", "profiling.usec_traj", "profiling.usec_ctrl", "profiling.usec_idle"]
  # genericLogTopic_log1_Variables: ["stabilizer.x", "ctrltarget.x", "vicon.x", "stabilizer.z", "ctrltarget.z", "vicon.z"]
  # genericLogTopic_log1_Variables: ["stateEstimate.roll", "ctrlMel.rolld"]
  genericLogTopic_log1_Variables: ["ctrlStat.edist"]
  # genericLogTopic_log1_Variables: ["stateEstimate.yaw", "ctrltarget.yaw", "stateEstimate.x", "ctrltarget.x"] #, "ctrltarget.x", "stateEstimate.roll", "ctrlMel.rolld"]
  firmwareParams: # for all (independent of type)
    commander:
      enHighLevel: 1
    stabilizer:
      estimator: 2 # 1: complementary, 2: kalman, 3: kalmanUSC (only crazyswarm firmware)
      controller: 1 # 1: PID, 2: mellinger
    ring:
      effect: 16 # 6: double spinner, 7: solid color, 16: packetRate
      solidBlue: 255 # if set to solid color
      solidGreen: 0 # if set to solid color
      solidRed: 0 # if set to solid color
      headlightEnable: 0
    #ekf:
    #  ext_var_xy: 1.5e-7 # 1e-7 # 1.5e-7
    #  ext_var_vel: 2e-4 # 2e-4
    #  ext_var_q: 4.5e-3 # 2e-3 # 4.5e-3
    #  gyro_var_xy: 0.2e-2 # 0.2e-4
    #  gyro_var_z: 0.2e-2 # 0.2e-4
    #  acc_var_xyz: 2.4e-3 # 2.4e-3
  # tracking
  motion_capture_type: "optitrack" # one of none,vicon,optitrack,qualisys,vrpn
  object_tracking_type: "motionCapture" # one of motionCapture,libobjecttracker
  # vicon_host_name: "vicon"
  optitrack_local_ip: "169.254.106.255"
  optitrack_server_ip: "169.254.41.1:3883"
  # qualisys_host_name: "10.0.5.219"
  # qualisys_base_port: 22222
  # vrpn_host_name: "vicon"
  # save_point_clouds: ~/pointCloud.ot
  print_latency: False
  write_csvs: False
  force_no_cache: False
  enable_parameters: True
  enable_logging: False

And then I start the launch file eric@eric-ThinkPad-S2:~$ roslaunch crazyswarm hover_swarm.launch

eric@eric-ThinkPad-S2:~$ roslaunch crazyswarm hover_swarm.launch ... logging to /home/eric/.ros/log/7b882882-5764-11e9-a618-34f64bc13130/roslaunch-eric-ThinkPad-S2-6968.log Checking log directory for disk usage. This may take awhile. Press Ctrl-C to interrupt Done checking log file disk usage. Usage is <1GB.

started roslaunch server http://eric-ThinkPad-S2:35681/

SUMMARY

PARAMETERS

NODES / crazyswarm_server (crazyswarm/crazyswarm_server) crazyswarm_teleop (crazyswarm/crazyswarm_teleop) joy (joy/joy_node) rviz (rviz/rviz)

ROS_MASTER_URI=http://localhost:11311

process[crazyswarm_server-1]: started with pid [6985] process[joy-2]: started with pid [6986] process[crazyswarmteleop-3]: started with pid [6987] process[rviz-4]: started with pid [6994] [ INFO] [1554442479.080369264]: Wait for services... [ INFO] [1554442479.095308661]: Manager ready. [ INFO] [1554442479.265023748]: Opened joystick: /dev/input/js0. deadzone: 0.050000. [crazyswarm_server-1] process has died [pid 6985, exit code -11, cmd /home/eric/crazyswarm/ros_ws/devel/lib/crazyswarm/crazyswarm_server __name:=crazyswarm_server __log:=/home/eric/.ros/log/7b882882-5764-11e9-a618-34f64bc13130/crazyswarm_server-1.log]. log file: /home/eric/.ros/log/7b882882-5764-11e9-a618-34f64bc13130/crazyswarm_server-1*.log

Everything is ok, but it goes wrong at last:[crazyswarm_server-1] process has died

What's wrong with it ? How to figure it out? Thanks!

whoenig commented 5 years ago

Looks like you are getting a segfault in the crazyswarm_server. To debug it, add "launch-prefix="xterm -e gdb --args"" to https://github.com/USC-ACTLab/crazyswarm/blob/master/ros_ws/src/crazyswarm/launch/hover_swarm.launch#L8. In the xterm window, type and enter "r" (for run), and after the crash occurred type and enter "bt" to get a stacktrace on where the crash occurred.

It's likely related to some wrong/missing configuration that does not have proper error checking.

lkyfly0765 commented 5 years ago

Hi whoenig, I have tried GDB debug process, and it says Screenshot from 2019-04-08 18-08-51 What's wrong with it ?

whoenig commented 5 years ago

Type "bt" and press enter to see a stack trace. If this does not contain line numbers, you'll need to recompile with debug symbols: "catkin_make clean" followed by "catkin_make -DCMAKE_BUILD_TYPE=Debug" in the ros_ws folder.

lkyfly0765 commented 5 years ago

Here is the stact trace. Screenshot from 2019-04-09 16-35-58

lkyfly0765 commented 5 years ago

Also I am doubt about the broadcast_address: "E7E7E7E7E7" what should the broadcast_address to be set? Every CF's radio address is not same.

whoenig commented 5 years ago

Which version/commit are you using? Those line numbers actually don't point to code in the latest master branch.

The broadcast address should not be changed from the default value (it's hard-coded in the firmware).

lkyfly0765 commented 5 years ago

I am using the latest version of the crazyswarm ros-package. We just change our package from Crazyflie_ros of your to crazyswarm because we are going to control more cfs. So I just clone the package at about a week a ago.
I will try to delete the package and clone it again.

I have changed the broadcast add to the “E7E7E7E7E7” in the launch file, will this cause something wrong ? Actually at the first time I use this launch file,the broadcast add had not been changed at that time, it also appears to the process has died.

Eric Liu

邮箱：lkyfly0765@126.com |

Signature is customized by Netease Mail Master

On 04/10/2019 08:44, whoenig wrote:

Which version/commit are you using? Those line numbers actually don't point to code in the latest master branch.

The broadcast address should not be changed from the default value (it's hard-coded in the firmware).

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub, or mute the thread.

whoenig commented 5 years ago

I couldn't find the lines even for the version from a week ago. You could re-clone, or attach the source code file (crazyswarm_server.cpp) here. If you change the broadcast address it should not crash (however, broadcast communication might not work correctly.) The crash is likely caused by some incorrect setting in crazyflieTypes.yaml or allCrazyflies.yaml that causes an out-of-bounds error.

lkyfly0765 commented 5 years ago

The allCrazyflies.yaml is: `crazyflies:

And the crazyflieTypes.yaml is not changed.Just the default configuration. Also we are facing another issue that when we use the python chooser.py to get the battery info of each cf, but in the xterm window it feedback as below(we have three cfs for test): eric@eric-ThinkPad-S2:~/crazyswarm/ros_ws/src/crazyswarm/scripts$ python chooser.py [rospack] Error: package 'crazyflie_tools' not found [rospack] Error: package 'crazyflie_tools' not found [rospack] Error: package 'crazyflie_tools' not found This problem occurs also when we try to reboot or sysoff the cfs. It feedback the same info [rospack] Error: package 'crazyflie_tools' not found. What caused this?

I am still testing the hover_swarm.launch file.So I am not sure whether the previous issue has been solved or not. Could you please help me to check my if my launch file is correct? I am using the Optitrack as the mocap system, and using the Motive 1.10 version. I select the markers and create the rigid body in Moitve so I choose motionCapture as the _object_trackingtype. Here I am confused that if the optitrack_local_ip: "localhost" should be change to my laptop IP or just remain default? I have a server computer who are running the Motive software and I use the Data-streaming to send the tracking info to my computer via cable. I attach my launch file below and could you please help me to check it ?

firmware: "crazyswarm" # one of "crazyswarm", "bitcraze" broadcast_address: "FFE7E7E7E7" world_frame: "/world" genericLogTopics: ["log1"] genericLogTopicFrequencies: [10] # genericLogTopic_log1_Variables: ["ekfprof.usec_setup", "ekfprof.usec_innov", "ekfprof.usec_gain", "ekfprof.usec_corr", "ekfprof.usec_cov"] # genericLogTopic_log1_Variables: ["profiling.usec_ekf", "profiling.usec_traj", "profiling.usec_ctrl", "profiling.usec_idle"] # genericLogTopic_log1_Variables: ["stabilizer.x", "ctrltarget.x", "vicon.x", "stabilizer.z", "ctrltarget.z", "vicon.z"] # genericLogTopic_log1_Variables: ["stateEstimate.roll", "ctrlMel.rolld"] genericLogTopic_log1_Variables: ["ctrlStat.edist"] # genericLogTopic_log1_Variables: ["stateEstimate.yaw", "ctrltarget.yaw", "stateEstimate.x", "ctrltarget.x"] #, "ctrltarget.x", "stateEstimate.roll", "ctrlMel.rolld"] firmwareParams: # for all (independent of type) commander: enHighLevel: 1 stabilizer: estimator: 3 # 1: complementary, 2: kalman, 3: kalmanUSC (only crazyswarm firmware) controller: 2 # 1: PID, 2: mellinger ring: effect: 16 # 6: double spinner, 7: solid color, 16: packetRate solidBlue: 255 # if set to solid color solidGreen: 0 # if set to solid color solidRed: 0 # if set to solid color headlightEnable: 0 #ekf: # ext_var_xy: 1.5e-7 # 1e-7 # 1.5e-7 # ext_var_vel: 2e-4 # 2e-4 # ext_var_q: 4.5e-3 # 2e-3 # 4.5e-3 # gyro_var_xy: 0.2e-2 # 0.2e-4 # gyro_var_z: 0.2e-2 # 0.2e-4 # acc_var_xyz: 2.4e-3 # 2.4e-3 # tracking motion_capture_type: "optitrack" # one of none,vicon,optitrack,qualisys,vrpn object_tracking_type: "motionCapture" # one of motionCapture,libobjecttracker # vicon_host_name: "vicon" optitrack_local_ip: "169.154.106.255" optitrack_server_ip: "169.254.41.1:3883" # qualisys_host_name: "10.0.5.219" # qualisys_base_port: 22222 # vrpn_host_name: "vicon" save_point_clouds: ~/pointCloud.ot print_latency: False write_csvs: False force_no_cache: False enable_parameters: True enable_logging: False
lkyfly0765 commented 5 years ago

I couldn't find the lines even for the version from a week ago. You could re-clone, or attach the source code file (crazyswarm_server.cpp) here. If you change the broadcast address it should not crash (however, broadcast communication might not work correctly.) The crash is likely caused by some incorrect setting in crazyflieTypes.yaml or allCrazyflies.yaml that causes an out-of-bounds error.

During my test, for Optitrack users, in hover_swarm.launch file L44 https://github.com/USC-ACTLab/crazyswarm/blob/master/ros_ws/src/crazyswarm/launch/hover_swarm.launch#L44, you have to configure the 'optitrack_local_ip:' to your ros laptop's IP add, or just comment out this line. If the IP add is not correct, it will occurs [crazyswarm_server-2] process has died when running the launch file. But I'm not sure will it connect to the Motive correctly if I comment it out.

lkyfly0765 commented 5 years ago

Here we find the solution of this problem. It's for Optitrack users. For the [crazyswarm-server-1] Process has died problems, please check the following options:

In hover_swarm.launch file L44 https://github.com/USC-ACTLab/crazyswarm/blob/master/ros_ws/src/crazyswarm/launch/hover_swarm.launch#L44, please confirm that the 'optitrack_local_ip:' to your ros laptop's IP address, and Don't comment this line or it will cause the connection problems of your computer with VRPN server.

For optitrack users, if you choose "motionCapture" as "object_tracking_type", you should create rigid body in Motive software.

But, IMPORTANTLY,Rigid Bodies must be created in order of cf2's ID number in allCrazyflies.yaml file.That is, cf1 must be created as the first one, and cf2 as the second one and cf3... A name of rigid body is also important. We highly recommended that you name the rigidbodies as cf1,cf2,cf3 and so on. However, its order is important! #reference https://github.com/USC-ACTLab/libmotioncapture/pull/3