In order to prevent transaction from waiting a long time to obtain a lock, use the lock_timeout parameter. It can be set globally or per transaction.

BEGIN;
SET LOCAL lock_timeout = '5s';
ALTER TABLE things ADD COLUMN foo_id INTEGER;
END;

This will timeout after 5 seconds if the transaction cannot obtain a lock. It will throw:

ERROR:  canceling statement due to lock timeout

This clause is used as part of a select query and can define:

• The lock strength. See also https://www.postgresql.org/docs/current/explicit-locking.html#LOCKING-ROWS
• Name the specific tables in the query whose rows should be locked. If no table is specified, all tables used in the query are locked.
• What action to take if encountering lock contention

See also https://www.postgresql.org/docs/current/sql-select.html#SQL-FOR-UPDATE-SHARE

These aren’t row locks, but worth mentioning. You can use these to control concurrency in general, even things not explicitly tied to data in the database. See also:

• https://www.postgresql.org/docs/current/explicit-locking.html#ADVISORY-LOCKS
• https://www.postgresql.org/docs/current/functions-admin.html#FUNCTIONS-ADVISORY-LOCKS

For using advisory locks in RoR, see https://github.com/ClosureTree/with_advisory_lock