Shutdown -r now halted VM

[cmdrrobin]

RancherOS Version: (ros os version)
version 1.1.0

Where are you running RancherOS? (docker-machine, AWS, GCE, baremetal, etc.)
ESX 5.5 (vSphere)

When you do a sudo shutdown -r now it halts the VM rather than a reboot.

[SvenDowideit]

the shutdown command is a really incomplete thing - I think it just ignores the -r

[gizmotronic]

I can confirm the same problem. As a workaround, sudo reboot works fine.

[vancluever]

Looking into this as part of #2121. Args are being ignored for sure, but it's specifically when the shutdown container is launched. Right now runDocker is not deriving its command arguments correctly, I have a fix which I will put in my PR as well.

You can see the specfiic power command go from LINUX_REBOOT_CMD_RESTART to LINUX_REBOOT_CMD_POWER_OFF in the extra log bits I put in below.

[            ] shutdown:info: power command is 0x1234567
[            ] shutdown:info: Setting /sbin/shutdown timeout to 60 (rancher.shutdown_timeout set to 60)
..�:[            ] shutdown:info: power command is 0x4321fedc
�b[            ] shutdown:info: Setting shutdown timeout to 60 (rancher.shutdown_timeout set to 60)
.�;[            ] shutdown:info: Stopping /ntp : 9a7ee89680bc
.�@.[            ] shutdown:info: Stopping /network : eb8309c55d8b
.�=.[            ] shutdown:info: Stopping /udev : e9d9d424e2c8
.�>.[            ] shutdown:info: Stopping /acpid : c70b4b5abbea
�>[            ] shutdown:info: Stopping /syslog : 0aa45d076f63
.�?.[            ] shutdown:info: Stopping /docker : b41522ed9f36
..........�M..........[            ] shutdown:info: Stopping /system-cron : 378f9eb375de
.[   59.991950] reboot: Power down

[janeczku]

Reproduced on v1.1.3 (VMware ESXi guest):
sudo shutdown -r now has inconsistent behaviour and results in machine powered off state instead of reboot 2 out of 5 times).
@niusmallnan