handlers: restart only the wireguard vpn updater supervisor process (fixes #47)

[DarshiniKurasa]

Summary

This PR improves deployment reliability by stopping Supervisor from restarting unrelated processes.
Replaces the global supervisorctl reload handler with a targeted restart of the WireGuard VPN updater supervisor process to avoid unnecessary restarts of unrelated services.

What was changed

• Added a new handler in handlers/main.yml:
  • Runs: supervisorctl restart openwisp-flask-vpn-updater-{{ openwisp2_wireguard_vpn_uuid }}
  • Uses register, failed_when, and changed_when logic so:
    • The play does not fail if the process doesn't exist
    • The task reports changed only when a restart succeeds
• Updated notify statements in tasks to use the new handler:
  • tasks/flask.yml
  • tasks/pip.yml
  • tasks/uwsgi.yml
• Kept verify.yml unchanged to avoid altering test behavior unnecessarily

Why

Reloading supervisor triggers a restart of all managed processes, creating avoidable downtime.
Restarting only the affected process is safer, more efficient, and aligns with the request in issue #47.

How to verify

Run local QA checks:

run-qa-checks

(Already tested locally, passed successfully)

Closes #47

Happy to update the PR if needed. Thanks for reviewing!