fix: avoid double restart of pmlogger on config changes

[kurik]

The pmlogger timers and control.d tasks notified a handler to restart pmlogger, but those same changes already set __pcp_restart_pmlogger and triggered a restart in the main service task. On first playbook run this could restart pmlogger twice in quick succession, causing systemd to fail the second start on Type=notify services.
Drop the redundant handler notifications and the Restart pmlogger handler; rely on the existing restart task, consistent with pmie.

Resolves: RHEL-182971

[coderabbitai]

📝 Walkthrough

Walkthrough

This change removes the pmlogger handler and consolidates restart logic into the task file. The handler removal eliminates the separate notification target, while task refactoring introduces targeted-host and single-control templating with conditional gates and a computed restart flag that unifies restart decisions based on detected changes.

Changes

pmlogger Restart Logic Refactoring

Layer / File(s)	Summary
Handler removal for pmlogger roles/pcp/handlers/main.yml	Deletes the Restart pmlogger handler task, leaving only handlers for pmcd, pmie, and pmproxy.
Task restart logic consolidation and config templating roles/pcp/tasks/pmlogger.yml	Removes direct restart notification from the retention/timers template task. Adds targeted-host templating for control.d layout per host and single-control templating task, each with conditional guards. Centralizes restart behavior by computing __pcp_restart_pmlogger from multiple change signals and choosing start vs restart based on boot status and that flag.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

Poem

🐰 Handlers fade, tasks take the wheel,
Pmlogger restarts now feel more real,
Control files dance in loops so neat,
Restart logic, centralized, complete! 🎯

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately describes the main change: avoiding double restart of pmlogger by removing redundant handler notifications.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description check	✅ Passed	The PR description clearly explains the issue, the fix, and includes a reference to the related Jira ticket, though it deviates from the template structure.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[kurik]

@richm may I ask you for review please? This is sort of a race condition I randomly observe during down stream testing on RHEL.