Skip to content

Crash when reallocating in rcl_clock_remove_jump_callback #938

@maxlein

Description

@maxlein

I don't really know how I can describe this best.
My robot is running 24/7. About 7 nodes running. (driver, laser, imu, localization, navigation, etc)

And somehow the localization_node from slam_toolbox is dying after ~1 to 3 hours, no other node.
It seems obvious that the toolbox does something wrong, but the stacktraces say something different.
-> toolbox issue
Crashes on executor, tf subscriber, message_filter ( issue )

Maybe someone of you guys has any idea what's going on here.
I am a bit clueless here.

Bug report

Required Info:

  • Operating System:
    • Ubuntu 18.04, Kernel 5.0, Core i5, 8GB RAM
  • Installation type:
    • binaries
  • Version or commit hash:
    • Eloquent
  • DDS implementation:
    • CycloneDDS
  • Client library (if applicable):
    • rclcpp

Steps to reproduce issue

Some setup for getting slam_toolbox working is needed.

I am running a "standard" robot stack.
No simulation, real hardware.

scan->base_link->odom->map
localization_node of slam_toolbox needs to run and after some time ( 1to 3 hours ) it crashes.

Expected behavior

No node crashes.

Actual behavior

localization_node crashes with different errors

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions