[systemd-devel] After upgrading glibc: Received environment initctl request

[Marti Raudsepp]

On Tue, Nov 1, 2011 at 19:26, Lennart Poettering <lennart at poettering.net> wrote:
> I didn't really give this much testing, so it would be good if you could
> check out if this works for you!

I built git version as of 263653e10353d8ad155f1faba01981816a2bb712 and
indeed this fixes the simple case.

> I have now commited a simple fix that will simply cause initctl to
> terminate itself after sending SIGTERM. If there's more data queued in
> the FIFO then systemd will respawn us anew, so things should be fine.

But this only solves half the problem. It doesn't cover the case when
you have systemd-initctl running, then issue "systemctl daemon-reexec"
manually and try run SysV /sbin/reboot.

Test case:
  telinit q        # spawn systemd-initctl
  systemctl daemon-reexec
  reboot        # doesn't do anything

syslog: systemd-initctl[349]: Failed to start unit: Did not receive a
reply. [...]

----
And there's a new problem now. When I manually start the initctl
service and it exits:
  systemctl start systemd-initctl.service
  telinit u

Then systemd thinks it's still running:
systemd-initctl.service - /dev/initctl Compatibility Daemon
	  Loaded: loaded (/lib/systemd/system/systemd-initctl.service; static)
	  Active: active (running) since Wed, 02 Nov 2011 18:43:21 +0200; 16s ago
	Main PID: 327
	  Status: "Shutting down..."
	  CGroup: name=systemd:/system/systemd-initctl.service

When trying to stop it, systemctl hangs for a long time, the service
is now in state:
	  Active: deactivating (stop-sigterm) since Wed, 02 Nov 2011 18:45:07
+0200; 39s ago

And eventually it goes into "failed" state after several minutes pass.

----
PS: Thanks to systemd, my VM now reboots in 1.5 seconds, so testing
this was pure pleasure. ;)

Regards,
Marti