We should go back through the items we had done in https://gitlab.com/gitlab-com/infrastructure/issues/4314 and make sure we scale up GCP hosts similar to what we had done in Azure as needed. * [x] API nodes * [x] Sidekiq workers * [x] Worker chef roles for sidekiq queues * [x] Any further tuning of NFS nodes * [x] Pgbouncer changes if any * [x] Prometheus? cc @andrewn