While I have identified a possible problem in the lxc start code (still to test), I note that there are several other unrelated defunct processes in some those VMs. For instance
lp 19479 19478 1 17:41 ? 00:00:00 [python3] <defunct> lp 19496 19478 6 17:41 ? 00:00:00 [python3] <defunct> utah 19590 19221 0 17:41 ? 00:00:00 sshd: utah@pts/4 utah 19592 19590 0 17:41 pts/4 00:00:00 -bash lp 19595 19478 0 17:41 ? 00:00:00 [foomatic] <defunct>
So I'm not sure there isn't either a problem with your host, or with the kernel (or with qemu for that matter).
While I have identified a possible problem in the lxc start code (still
to test), I note that there are several other unrelated defunct
processes in some those VMs. For instance
lp 19479 19478 1 17:41 ? 00:00:00 [python3] <defunct>
lp 19496 19478 6 17:41 ? 00:00:00 [python3] <defunct>
utah 19590 19221 0 17:41 ? 00:00:00 sshd: utah@pts/4
utah 19592 19590 0 17:41 pts/4 00:00:00 -bash
lp 19595 19478 0 17:41 ? 00:00:00 [foomatic] <defunct>
So I'm not sure there isn't either a problem with your host, or with
the kernel (or with qemu for that matter).