Bug 1296123

service skyringd start
service skyringd status
service skyringd stop

Using rhscon-core-0.0.8-2.el7.x86_64, I see that skyringd service is available:

~~~
# systemctl list-unit-files | grep skyring
skyringd.service                            static
# systemctl status skyringd
● skyringd.service - skyring server
   Loaded: loaded (/usr/lib/systemd/system/skyringd.service; static; vendor preset: disabled)
   Active: active (running) since Fri 2016-02-12 13:51:03 CET; 6min ago
  Process: 31668 ExecStart=/etc/init.d/skyringd start (code=exited, status=0/SUCCESS)
 Main PID: 31673 (skyring)
   CGroup: /system.slice/skyringd.service
           ├─31673 /usr/bin/skyring
           └─31693 /var/lib/skyring/providers/bigfin {"config":{"host":"0.0.0.0","httpPort":8080,"supportedversions":[1]},"logging":{"logToStderr":false,"filename":"/var/log/skyring/skyring.log"...

Feb 12 13:51:03 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Starting skyring server...
Feb 12 13:51:03 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Started skyring server.
Feb 12 13:51:03 mbukatov-usm1-server.os1.phx2.redhat.com skyringd[31668]: Starting skyring:

~~~

That said, I see a big problem with the way it has been implemented:

Systemd service file is just a wrapper. It doesn't start skyring process
directly, but it calls system v style init script which then starts the
skyring process, see:

~~~
# systemctl cat skyringd
# /usr/lib/systemd/system/skyringd.service
[Unit]
Description=skyring server

[Service]
Type=forking
KillMode=process
ExecStart=/etc/init.d/skyringd start
ExecStop=/etc/init.d/skyringd stop
# file /etc/init.d/skyringd
/etc/init.d/skyringd: POSIX shell script, ASCII text executable
# wc -l /etc/init.d/skyringd
105 /etc/init.d/skyringd
~~~

Such approach is not acceptable for a brand new service. Moreover no systemd
features could be configured/used this way.

Eg. instead of hacking handling of pid and lock files in a shell init script, 
you can specify it in a systemd service file.

This direclty leads to several problems, such as (note that the list below
doesn't cover all possible problems):

1) systemd is not avare of any skyring socket:

~~~
# systemctl list-sockets --all | grep skyring
#
~~~

note: compare this with sshd:

~~~
# systemctl list-sockets --all | grep ssh
[::]:22                         sshd.socket
~~~

2) systemd is not aware of skyring service dependencies:

~~~
# systemctl list-dependencies skyringd | grep -i mongo
#
~~~

note: I would expect to see mongob here 

3) When I ask journalctl for logs of skyringd service, I see just list of 
start and stop events. I would expect to see more details in the journal
logs. Eg. I started a create cluster task and still see nothing in the logs.

~~~
# journalctl -u skyringd
-- Logs begin at Tue 2016-02-09 12:56:17 CET, end at Fri 2016-02-12 14:03:43 CET. --
Feb 09 13:35:59 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Starting skyring server...
Feb 09 13:35:59 mbukatov-usm1-server.os1.phx2.redhat.com skyringd[26966]: Starting skyring:
Feb 09 13:35:59 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Started skyring server.
Feb 09 13:36:00 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Started skyring server.
Feb 12 13:50:50 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Stopping skyring server...
Feb 12 13:50:52 mbukatov-usm1-server.os1.phx2.redhat.com skyringd[31602]: Stopping skyring: [  OK  ]
Feb 12 13:50:52 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Stopped skyring server.
Feb 12 13:51:03 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Starting skyring server...
Feb 12 13:51:03 mbukatov-usm1-server.os1.phx2.redhat.com systemd[1]: Started skyring server.
Feb 12 13:51:03 mbukatov-usm1-server.os1.phx2.redhat.com skyringd[31668]: Starting skyring:
~~~

Moreover: looking at /usr/bin/skyring-setup.sh script, it's obvious that some
of the stuff here should be moved into systemd service file (eg. list of services
required by skyring: salt-master, influxdb and mongod.

Also if there is proper systemd service file this will not happen:
2016-02-19T12:28:16.782+01:00 ERROR    ldapauth.go:366 AddUser] User admin already exists
2016-02-19T12:28:16.782+01:00 ERROR    auth.go:37 AddDefaultUser] Unable to create default User:user already exists
2016-02-19T12:28:16.785+01:00 INFO     skyring.go:500 InitializeApplication] Starting clusters syncing
2016-02-19T12:28:16.785+01:00 INFO     main.go:187 start] Starting the providers
2016-02-19T12:28:16.785+01:00 DEBUG    provider-conf.go:53 LoadProviderConfig] File Name: ceph.conf
2016-02-19T12:28:16.786+01:00 DEBUG    provider-conf.go:68 LoadProviderConfig] Collection:%!(EXTRA []conf.ProviderInfo=[{{ceph bigfin} []}])
2016-02-19T12:28:16.786+01:00 INFO     skyring.go:92 StartProviders] Config:%!(EXTRA []conf.ProviderInfo=[{{ceph bigfin} []}])
2016-02-19T12:28:16.786+01:00 DEBUG    skyring.go:95 StartProviders] Config:%!(EXTRA conf.ProviderInfo={{ceph bigfin} []})
2016-02-19T12:28:16.798+01:00 INFO     main.go:194 start] Starting event listener
2016-02-19T12:28:16.801+01:00 CRITICAL listener.go:134 StartListener] listen error:listen unix /var/run/.skyring-event: bind: address already in use

Also note that with the current service file it's not possible to enable the
service. Which means that skyring service would not be running after reboot,
and a manual start is required.

~~~
# systemctl enable skyringd.service
The unit files have no [Install] section. They are not meant to be enabled
using systemctl.
Possible reasons for having this kind of units are:
1) A unit may be statically enabled by being symlinked from another unit's
   .wants/ or .requires/ directory.
2) A unit's purpose may be to act as a helper for some other unit which has
   a requirement dependency on it.
3) A unit may be started when needed via activation (socket, path, timer,
   D-Bus, udev, scripted systemctl call, ...).
# systemctl status skyringd.service
● skyringd.service - skyring server
   Loaded: loaded (/usr/lib/systemd/system/skyringd.service; static; vendor preset: disabled)
   Active: active (running) since Fri 2016-03-18 16:06:09 CET; 4min 52s ago
 Main PID: 3610 (skyring)
   CGroup: /system.slice/skyringd.service
           ├─3610 /usr/bin/skyring
           └─3631 /var/lib/skyring/providers/bigfin {"config":{"host":"0.0.0.0","httpPort":8080,"supportedversions":[1]},"logging":{"logToStderr":false,"fi...
#
~~~

Checking rhscon-core-0.0.41-1.el7scon.x86_64

Skyring now has a proper systemd service unit file:

* start, stop and restart is possible
* enable/disable of the service works
* both mongod and salt-master services are specified as dependencies
  (starting skyring would start them if they are not running)

Moreover, I'm able to check logs via journal using:

~~~
# journalctl -u skyring
~~~