Question: How is it possible to end up with defunct proceses using async_process::Command with the default reap_on_drop?

[boombatower]

Presumably, I am missing something, but I haven't been able to put my finger on it. During testing and intentionally trying to not await spawn()'d processes I have been unable to create zombie processes, but during long-running tests I eventually get defunct rsync processes.

Everything is immediately await'd and I even tried not awaying and just dropping the variable. I am not seeing how I end up with defunct processes.

Using ps -ejf I see my main process PID 1 since in a container and then PPID of all the defunct rsync processes of 1. The main processes spawns thousands of rsync calls over the course of hours...almost all of which work and are cleaned up, but for some reason I end up with some that don't.

UID         PID   PPID   PGID    SID  C STIME TTY          TIME CMD
root          1      0      1      1  0  2024 ?        00:01:28 /my-rust-app
root        163      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        164      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        256      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        257      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        302      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        303      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        423      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root        424      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1025      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1026      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1258      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1259      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1276      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1277      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1296      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>
root       1297      1      1      1  0  2024 ?        00:00:00 [rsync] <defunct>

What scenario would need to occur to cause this and perhaps I can sort out how I am pulling that off.

Command::new("/usr/bin/rsync")
  .args([...])
  .current_dir(...)
  .env_clear()
  .kill_on_drop(true)
  .spawn()