Some web APIs stream database results through the web API without first buffering them in an array. Examples, are the web APIs for querying instances, holdings, and items in the Inventory module and the users web API in the users module. It appears that /records doesn't do this. If you don't limit the results to something like 1,000 and page through the data, the server process will crash with an out of memory error, or, Docker will kill off the process. One should be able to query and get back all records without having to page through the data.