Express Create Migration Pgdump

For the user to run — the skill does not execute these commands.

1. From a machine with PostgreSQL client tooling and network access to both clusters, dump the Express cluster:

   Copy

   pg_dump \
     --host <express-cluster-endpoint> \
     --port 5432 \
     --username <master-user> \
     --dbname <database> \
     --format=custom \
     --file <database>.dump

2. Restore into the Full Configuration cluster:

   Copy

   pg_restore \
     --host <full-config-cluster-endpoint> \
     --port 5432 \
     --username <master-user> \
     --dbname <database> \
     <database>.dump

Illustrative only. Adjust flags: --no-owner, --no-privileges, --clean, --create, --jobs N for parallel restore.

Credentials: retrieve the password from AWS Secrets Manager at run time and pass it to the client via a temporary ~/.pgpass file (chmod 600, deleted after) referenced by PGPASSFILE — do NOT use export PGPASSWORD (visible in the process environment via /proc/<pid>/environ) or inline --password. Better still, if the source cluster has IAM database authentication enabled, generate a short-lived token with aws rds generate-db-auth-token and use that instead of a long-lived password. Source: PostgreSQL pg_dump (opens in a new tab) and pg_restore (opens in a new tab) docs.

Considerations:

• The dump is a logical export; extensions, roles, and ownership metadata may need special handling. Use --no-owner and --no-privileges if the target has different role names.
• Large objects and sequences may need explicit handling.
• Downtime is dump + restore time, scaling roughly linearly with data size. For anything larger than a dev dataset, Path 1 or Path 2 is usually better.