Vmtoolsd not stopping

Question

Hello,
I'm playing around with Arch Linux in a VMware Virtual Machine, preparing to switch from Ubuntu which I was using the last 7 years. So far everything went pretty well, but now I have a problem which I can't solve by myself.
Since the last pacman -Syu vmtoolsd refuses to stop, when shutting down. After 90 seconds, systemd kills it and the shutdown continues.
Nov 03 07:04:22 vm-arch systemd[1]: Stopped Permit User Sessions.
Nov 03 07:04:22 vm-arch systemd[1]: Stopped dhcpcd on eth0.
Nov 03 07:04:22 vm-arch systemd[1]: Unit [email protected] entered failed state
Nov 03 07:04:22 vm-arch systemd[1]: Stopping Remote File Systems.
Nov 03 07:04:22 vm-arch systemd[1]: Stopped target Remote File Systems.
Nov 03 07:05:52 vm-arch systemd[1]: vmtoolsd.service stopping timed out. Killing.
Nov 03 07:05:52 vm-arch systemd[1]: vmtoolsd.service: main process exited, code=killed, status=9/KILL
Nov 03 07:05:52 vm-arch systemd[1]: Stopped Open Virtual Machine Tools (VMware Tools).
Nov 03 07:05:52 vm-arch systemd[1]: Unit vmtoolsd.service entered failed state
Nov 03 07:05:52 vm-arch systemd[1]: Stopping Basic System.
Nov 03 07:05:52 vm-arch systemd[1]: Stopped target Basic System.
Does anyone else have this problem? How can I find out more about what's going wrong?

Answer 1

alexoz wrote:
@OP: I have the same issue. It looks like vmtoolsd does not exit when a SIGTERM is received so systemd waits for 90 seconds before sending a SIGKILL. I solved this by editing /usr/lib/systemd/system/vmtoolsd.service and adding a KillSignal=SIGKILL line in the [Service] section. This way systemd sends a SIGKILL instead of a SIGTERM.
Another way to overcome this issue would be to set the timeout for this service to just a few seconds instead of the default 90, by adding e.g. TimeoutStopSec=2 (2 seconds) in the [Service] section.
Thanks for the tip. I really should take the time to learn systemd service syntax.
I did an strace on vmtoolsd, and it seems to get stuck waiting for some lock:
--- SIGTERM {si_signo=SIGTERM, si_code=SI_USER, si_pid=2734, si_uid=1000} ---
write(13, "\17\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\256\n\0\0\350\3\0\0\200\373\2\1\0\0\0\0"..., 128) = 128
rt_sigreturn() = -1 EINTR (Interrupted system call)
poll([{fd=3, events=POLLIN}, {fd=12, events=POLLIN}, {fd=12, events=POLLIN}, {fd=12, events=POLLIN}, {fd=12, events=POLLIN}, {fd=12, events=POLLIN}], 6, 732) = 5 ([{fd=12, revents=POLLIN}, {fd=12, revents=POLLIN}, {fd=12, revents=POLLIN}, {fd=12, revents=POLLIN}, {fd=12, revents=POLLIN}])
read(12, "\17\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\256\n\0\0\350\3\0\0\200\373\2\1\0\0\0\0"..., 128) = 128
write(3, "\1\0\0\0\0\0\0\0", 8) = 8
write(3, "\1\0\0\0\0\0\0\0", 8) = 8
write(3, "\1\0\0\0\0\0\0\0", 8) = 8
write(3, "\1\0\0\0\0\0\0\0", 8) = 8
write(3, "\1\0\0\0\0\0\0\0", 8) = 8
futex(0x100fe00, FUTEX_WAKE_PRIVATE, 1) = 0
futex(0x100fe00, FUTEX_WAIT_PRIVATE, 2, NULL <unfinished ...>
+++ killed by SIGKILL +++
Killed
Someone reported it upstream but the developers didn't bother to check for the bug and just dismissed it with "we don't support systemd". And it doesn't bother me enough to try and fix it myself.