autopkgtests failing with systemd-232

The autopkgtests for dogtag-pki are failing. It looks like this started with the upgrade of systemd to 232.

Previously, pki-tomcatd was marked as failed on startup:

Job for pki-tomcatd.service failed because the control process exited with error code.
See "systemctl status pki-tomcatd.service" and "journalctl -xe" for details.
invoke-rc.d: initscript pki-tomcatd, action "start" failed.
● pki-tomcatd.service - LSB: Start pki-tomcatd at boot time
   Loaded: loaded (/etc/init.d/pki-tomcatd; generated; vendor preset: enabled)
   Active: failed (Result: exit-code) since Mon 2016-11-07 20:51:19 UTC; 14ms ago
     Docs: man:systemd-sysv-generator(8)
  Process: 8100 ExecStart=/etc/init.d/pki-tomcatd start (code=exited, status=5)

Now, the service is marked as started and exited:

● pki-tomcatd.service - LSB: Start pki-tomcatd at boot time
   Loaded: loaded (/etc/init.d/pki-tomcatd; generated; vendor preset: enabled)
   Active: active (exited) since Tue 2017-02-14 06:02:25 UTC; 31s ago
     Docs: man:systemd-sysv-generator(8)

Since systemd-sysv-generator uses RemainAfterExit=true, subsequent "systemctl start pki-tomcatd" invocations do nothing.

I believe the relevant systemd change is:

https://github.com/systemd/systemd/commit/41e2036eb83204df95a1c3e829bcfd78ee17aaa3

which fixed it to detect the special LSB exit codes as intended.

I see that .../scriptlets/configuration.py issues start() when configuring the first tomcat instance and restart() for subsequent instances (line 364). Maybe one workaround would be to use restart() unconditionally for now? That looks like it does roughly the right thing.