gazebosim / gz-sim

Open source robotics simulator. The latest version of Gazebo.
https://gazebosim.org
Apache License 2.0
732 stars 273 forks source link

LogSystemTest.RecordAndPlayback fails in ign-gazebo4 #435

Open j-rivero opened 4 years ago

j-rivero commented 4 years ago

The pull #410 was merge with failures in Github actions.

 [ RUN      ] LogSystemTest.RecordAndPlayback
  [Msg] Loading SDF world file[/github/workspace/test/worlds/log_record_dbl_pendulum.sdf].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [2251689575469537287] / [ign_gazebo_components.World].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [13994732549916512682] / [ign_gazebo_components.Name].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [12592746352568925681] / [ign_gazebo_components.Gravity].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [13224937992534617849] / [ign_gazebo_components.MagneticField].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [3630648173860223239] / [ign_gazebo_components.PhysicsEnginePlugin].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [17459188283658606303] / [ign_gazebo_components.RenderEngineServerPlugin].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [8705992680619689917] / [ign_gazebo_components.RenderEngineGuiPlugin].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [8753193699724811771] / [ign_gazebo_components.Wind].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [12173050716021724529] / [ign_gazebo_components.WorldLinearVelocity].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [15943768124495574352] / [ign_gazebo_components.WorldLinearVelocitySeed].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [3297509811873971798] / [ign_gazebo_components.ParentEntity].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [17100615127981600159] / [ign_gazebo_components.Scene].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [17605309075052480649] / [ign_gazebo_components.Atmosphere].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [8064491505919932473] / [ign_gazebo_components.Level].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [2668898242563798256] / [ign_gazebo_components.DefaultLevel].
  [Dbg] [EntityComponentManager.cc:661] Using components of type [11371360182141354106] / [ign_gazebo_components.LevelEntityNames].
  [Dbg] [Physics.cc:620] Loaded [ignition::physics::dartsim::Plugin] from library [/usr/lib/x86_64-linux-gnu/ign-physics-3/engine-plugins/libignition-physics-dartsim-plugin.so]
  [Dbg] [SimulationRunner.cc:765] Loaded system [ignition::gazebo::systems::Physics] for entity [1]

...

  [Msg] Found no publishers on /stats, adding root stats topic
  [Msg] Found no publishers on /clock, adding root clock topic
  [Dbg] [SimulationRunner.cc:403] Creating PostUpdate worker threads: 2
  [Dbg] [SimulationRunner.cc:416] Creating postupdate worker thread (0)
  [Msg] Serving scene information on [/world/default/scene/info]
  [Msg] Serving graph information on [/world/default/scene/graph]
  [Msg] Serving full state on [/world/default/state]
  [Msg] Publishing scene information on [/world/default/scene/info]
  [Msg] Publishing entity deletions on [/world/default/scene/deletion]
  [Msg] Publishing state changes on [/world/default/state]
  [Msg] Publishing pose messages on [/world/default/pose/info]
  [Msg] Publishing dynamic pose messages on [/world/default/dynamic_pose/info]
  /github/workspace/test/integration/log_system.cc:895: Failure
  Value of: nTotal == expectedPoseCount || nTotal == expectedPoseCount + 1
    Actual: false
  Expected: true
  nTotal [32] expectedPoseCount [30]
  [Dbg] [SimulationRunner.cc:432] Exiting postupdate worker thread (0)
  [  FAILED  ] LogSystemTest.RecordAndPlayback (1960 ms)

Seems to me like a possible timing/race condition problem. The test is disable for APPLE, not sure if this is related.

chapulina commented 4 years ago

Seems to me like a possible timing/race condition problem.

Yeah this test has been flaky. We added the || condition to relax it a bit and it still fails sometimes. I'll see what I can do.