Some background jobs need to be manually re-run after using the postgresql port script

This issue has been migrated from [#9758](https://github.com/matrix-org/synapse/issues/9758).

---

### Description

I followed the steps in https://github.com/matrix-org/synapse/blob/develop/docs/postgres.md in order to port synapse db from sqlite3 to postgresql. After having finished I couldn't find existing users via the direct chat dialog ("search_all_users" was true).
Maybe the port script is broken?

### Steps to reproduce

- follow the porting steps in the link above
- try searching for users

I would expect to find users after porting to postgresql.

### Version information

- **Homeserver**:

If not matrix.org:

- **Version**: "server_version":"1.30.1","python_version":"3.8.5"

- **Platform**: Debian

### Workaround

Thanks to Aaron I could get this to work via:

INSERT INTO background_updates (update_name, progress_json) VALUES
('populate_user_directory_createtables', '{}');

-- Run through each room and update the user directory according to who is in it
INSERT INTO background_updates (update_name, progress_json, depends_on) VALUES
('populate_user_directory_process_rooms', '{}', 'populate_user_directory_createtables');

-- Insert all users, if search_all_users is on
INSERT INTO background_updates (update_name, progress_json, depends_on) VALUES
('populate_user_directory_process_users', '{}', 'populate_user_directory_process_rooms');

-- Clean up staging tables
INSERT INTO background_updates (update_name, progress_json, depends_on) VALUES
('populate_user_directory_cleanup', '{}', 'populate_user_directory_process_users');