Design how scale code matching on different machines

Some options includes:

- [ ] Co-locating matchcode on the purldb server could be very taxing when many requests are happening
- [ ] Consider keeping matchcode on a different machine and access the purldb database remotely
- [ ] Consider creating a purldb database replica on a different server and use that for matching
- [ ] In the case where we have matchcode on a separate machine than purldb, we can have matchcode on its own host with the single matching endpoint exposed and we could avoid the use of traefik
- [ ] Look into database replication with a secure remote access