Restart Foreman services after package upgrade on Debian

[ehelms]

This puts Debian on the same footing as RPM where we do not start
services on initial install but we do restart them after package
upgrades.

Opening as a draft so I can get some builds to test.

[ekohl]

If this does solve things, I'd appreciate a Redmine issue.

[ehelms]

Forgot I needed changelog updates to trigger builds, but I am feeling like I will need #7211 and the CI change to really get testable builds.

Edit: and I think the current failure agrees with me

[ehelms]

[test deb]

[ehelms]

[test deb]

[ehelms]

[test deb]

[ehelms]

[test deb]

[ehelms]

I was able to complete testing of this change and can confirm this resolves the problem. You can see that foreman-service gets updated:

2021-11-02 17:58:33 configure foreman-service:all 9999-3.1.0-buster+scratchbuild+20211102165550 <none>
2021-11-02 17:58:33 status unpacked foreman-service:all 9999-3.1.0-buster+scratchbuild+20211102165550
2021-11-02 17:58:33 status half-configured foreman-service:all 9999-3.1.0-buster+scratchbuild+20211102165550
2021-11-02 17:59:00 status installed foreman-service:all 9999-3.1.0-buster+scratchbuild+20211102165550

And the service gets restarted:

Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 systemd[1]: Stopping Foreman...
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 foreman[22245]: [22245] === puma shutdown: 2021-11-02 17:58:50 +0000 ===
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 foreman[22245]: [22245] - Goodbye!
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 foreman[22245]: [22245] - Gracefully shutting down workers...
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 foreman[22245]: Exiting
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 systemd[1]: foreman.service: Main process exited, code=killed, status=15/TERM
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 systemd[1]: foreman.service: Succeeded.
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 systemd[1]: Stopped Foreman.
Nov 02 17:58:50 pipe-up-foreman-nightly-debian10 systemd[1]: Starting Foreman...
Nov 02 17:58:52 pipe-up-foreman-nightly-debian10 foreman[22369]: => Booting Puma
Nov 02 17:58:52 pipe-up-foreman-nightly-debian10 foreman[22369]: => Rails 6.0.3.7 application starting in production
Nov 02 17:58:52 pipe-up-foreman-nightly-debian10 foreman[22369]: => Run `rails server --help` for more startup options
Nov 02 17:58:53 pipe-up-foreman-nightly-debian10 foreman[22369]: /usr/share/foreman/vendor/ruby/2.5.0/gems/foreman_puppet-2.0.0.alpha.2/lib/foreman_puppet/register.rb:120: warning: already initialized constant Foreman::Plugin::RbacSupport::AUTO_EXTENDED_ROLES
Nov 02 17:58:53 pipe-up-foreman-nightly-debian10 foreman[22369]: /usr/share/foreman/app/registries/foreman/plugin/rbac_support.rb:5: warning: previous definition of AUTO_EXTENDED_ROLES was here
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] * Enabling systemd notification integration
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] Puma starting in cluster mode...
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] * Puma version: 5.5.2 (ruby 2.5.5-p157) ("Zawgyi")
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] *  Min threads: 5
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] *  Max threads: 5
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] *  Environment: production
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] *   Master PID: 22369
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] *      Workers: 2
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] *     Restarts: (✔) hot (✖) phased
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] * Preloading application
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] * Activated unix:///run/foreman.sock
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] Use Ctrl-C to stop
Nov 02 17:58:59 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] * Starting control server on unix:///usr/share/foreman/tmp/sockets/pumactl.sock
Nov 02 17:59:00 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] - Worker 1 (PID: 22405) booted in 0.21s, phase: 0
Nov 02 17:59:00 pipe-up-foreman-nightly-debian10 foreman[22369]: [22369] - Worker 0 (PID: 22400) booted in 0.22s, phase: 0
Nov 02 17:59:00 pipe-up-foreman-nightly-debian10 systemd[1]: Started Foreman.

[evgeni]

nice this got actually built as foreman-3.1.0-2, not with nightly versioning.

fix in theforeman/jenkins-jobs#132

[ekohl]

Generally we don't keep a changelog for all changes within the release cycle. We do end up with this weird bit where we sort of go back in time.

foreman-packaging/debian/bionic/foreman/changelog

Lines 13 to 35 in 2639d9e

	foreman (3.0.0-1) stable; urgency=low
	* 3.0.0 released
	-- Patrick Creech <pcreech@redhat.com> Wed, 08 Sep 2021 09:42:36 -0400
	foreman (3.0.0~rc2-1) stable; urgency=low
	* 3.0.0-rc2 released
	-- Patrick Creech <pcreech@redhat.com> Wed, 25 Aug 2021 14:32:40 -0400
	foreman (3.0.0~rc1-1) stable; urgency=low
	* 3.0.0-rc1 released
	-- Ewoud Kohl van Wijngaarden <ewoud@kohlvanwijngaarden.nl> Fri, 06 Aug 2021 17:17:44 +0200
	foreman (3.0.0-1) stable; urgency=low
	* Bump changelog to 3.0.0 to match VERSION
	-- Tomer Brisker <tbrisker@gmail.com> Thu, 22 Jul 2021 17:02:34 +0300

I wonder if having this newer version in the changelog can cause problems later on.

[evgeni]

Nope, while not nice, nothing in the Debian stack looks at older changelog entries.

[ekohl]

I wonder if instead we should "merge" the changelog entries

[evgeni]

For sake of cleanup? Maybe.