Here's status.  It looks about the same other than that benign log entry about the PID

# systemctl status backuppc

● backuppc.service - BackupPC server

   Loaded: loaded (/etc/systemd/system/backuppc.service; enabled)

   Active: active (running) since Wed 2017-04-12 18:13:58 UTC; 7s ago

  Process: 4325 ExecStart=/usr/local/BackupPC/bin/BackupPC -d (code=exited, status=0/SUCCESS)

 Main PID: 4328 (BackupPC)

   CGroup: /system.slice/backuppc.service

           ├─4328 /usr/bin/perl /usr/local/BackupPC/bin/BackupPC -d

           ├─4329 /usr/bin/perl /usr/local/BackupPC/bin/BackupPC_Admin_SCGI

           └─4330 /usr/bin/perl /usr/local/BackupPC/bin/BackupPC_Admin_SCGI

Apr 12 18:13:58 us-backuppc systemd[1]: Starting BackupPC server...

Apr 12 18:13:58 us-backuppc systemd[1]: PID file /var/run/BackupPC/BackupPC.pid not readable (yet?) after start.

Apr 12 18:13:58 us-backuppc systemd[1]: Started BackupPC server.

On Wed, Apr 12, 2017 at 1:13 PM Richard Shaw <hobbes1069 AT gmail DOT com> wrote:

On Wed, Apr 12, 2017 at 1:06 PM, Nicholas Hall <ngharo AT gmail DOT com> wrote:

Good question!  

I have an update: everything works fine when started via systemd when I change the unit file to Type=forking.  Wondering how reproducible the problem is on other platforms...

I was going to ask about that but said, "nah, that can't be it"...

Can you do a systemctl status while running with forking? Are additional processes spawned?

Thanks,

Richard