Close PG connection when Client is dropped

[martykulma]

Reintroduces #33322 and changes how PG replication slots are managed.

#33322 exposed a race in slot creation that could result in the permanent slot being visible just long enough to create a temporary slot, and then disappear. This resulted in PG progressing its LSN and MZ detecting the slot was over compacted.

• replication slots are only created by the snapshot operator
  • permanent replication slot (NOEXPORT_SNAPSHOT) is always created
  • temporary replication slot is always used for snapshot, this only gets created if a snapshot is needed
• replication operator no longer creates a slot, but still checks for slot existence and terminates any active PID using the slot

Motivation

fixes https://github.com/MaterializeInc/database-issues/issues/7288

Checklist

• This PR has adequate test coverage / QA involvement has been duly considered. (trigger-ci for additional test/nightly runs)
• This PR has an associated up-to-date design doc, is a design doc (template), or is sufficiently small to not require a design.
• If this PR evolves an existing $T ⇔ Proto$T mapping (possibly in a backwards-incompatible way), then it is tagged with a T-proto label.
• If this PR will require changes to cloud orchestration or tests, there is a companion cloud PR to account for those changes that is tagged with the release-blocker label (example).
• If this PR includes major user-facing behavior changes, I have pinged the relevant PM to schedule a changelog post.

[martykulma]

Rerunning the platform checks in nightly (ran once already, but I've rebased since to pull in removing yugabyte).