2. To the best of my knowledge, Primal has a backend service called 'caching-service.' This service collects and stores events sent by predefined relays, so that Primal clients can retrieve information from the caching service without having to connect to each relay individually. The Github project provides a list of relays for the caching-service, but which ones are connected in the Production environment is likely known only to the operations team of Primal. This is my speculation based on the Github project provided by Primal and some technical architecture descriptions.