fix(postgres-config): always restart if --no-restart is omitted

[avallete]

What kind of change does this PR introduce?

With the new api changes: https://github.com/supabase/infrastructure/pull/20462

We must always the restart_database setting to the endpoint. Otherwise we go from the logic of "always restart except if --no-restart is specified" to a "restart only if the changed settings require a restart and --no-restart is not specified".

Which in the case of maintenance_work_mem might sometimes don't work as expected

Related: https://app.hubspot.com/live-messages/19953346/inbox/8348874005#comment

PS: An alternative would be to mark maintenance_work_mem as a parameter requiring restart on api side, but according to this documentation: https://postgresqlco.nf/doc/en/param/maintenance_work_mem/ it isn't the case.

What is the current behavior?

Please link any relevant issues here.

What is the new behavior?

Feel free to include screenshots if it includes visual changes.

Additional context

Add any other context or screenshots.

[coveralls]

Pull Request Test Coverage Report for Build 11795862893

Details

• 0 of 2 (0.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.03%) to 59.77%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
internal/postgresConfig/update/update.go	0	2	0.0%

Totals
Change from base Build 11794104619:	0.03%
Covered Lines:	6381
Relevant Lines:	10676

---

💛 - Coveralls

[sweatybridge]

Which in the case of maintenance_work_mem might sometimes don't work as expected

I think a better fix for this is on infra worker, by sending a post request to reload config to admin api running on both primary and read replicas. This is needed whenever we update config without restarting database. More context https://supabase.slack.com/archives/C02FHG9QQAF/p1729922433212149?thread_ts=1728485737.099679&cid=C02FHG9QQAF

[avallete]

Which in the case of maintenance_work_mem might sometimes don't work as expected

I think a better fix for this is on infra worker, by sending a post request to reload config to admin api running on both primary and read replicas. This is needed whenever we update config without restarting database. More context https://supabase.slack.com/archives/C02FHG9QQAF/p1729922433212149?thread_ts=1728485737.099679&cid=C02FHG9QQAF

Closing in favor of: https://github.com/supabase/infrastructure/pull/20575