fix(postgres): add default StopGracePeriod to prevent WAL corruption

[OthmanHaba]

What is this PR about?

Please describe in a short paragraph what this PR is about.

Summary

Adds a default 30-second StopGracePeriod for PostgreSQL Swarm services to prevent WAL corruption on redeployment and service stop/restart

Issues related

#3595

Problem

When redeploying or stopping a PostgreSQL service (especially on external servers), Docker Swarm sends SIGTERM and then SIGKILL after the grace period expires. Previously, StopGracePeriod was only applied if the user explicitly configured it — otherwise Docker's default of 10 seconds was used.

10 seconds is often not enough for PostgreSQL to complete its shutdown sequence (flush WAL buffers, write a final checkpoint). If SIGKILL arrives before that finishes, the WAL is left in an inconsistent state, causing this on the next startup:

PANIC: could not locate a valid checkpoint record
This makes the database unrecoverable without manual intervention (pg_resetwal).

Fix

Changed StopGracePeriod in buildPostgres() from opt-in to always-present, with a default fallback of 30 seconds (30,000,000,000 nanoseconds). If the user has configured a custom value, that value is still respected.

• ...(StopGracePeriod !== null &&

• StopGracePeriod !== undefined && { StopGracePeriod }),
  

• StopGracePeriod: StopGracePeriod ?? 30_000_000_000,
  closes Postgres Template Database corrupted after assigning external port #3595