better error message when schema service memory limit is reached #7191
Description
When sending a huge service schema (10mb; see attached file) to our GraphQL API, the schema service will attempt composition in a worker and kill the worker due to exceeding the memory limit. In that case we raise an unexpected error to the enduser. Instead we want the CLI user to see something like "Your schema is too big, please reach out to us". So in that case we can investigate whether it is a memory leak/infinite loop or the customer actually has a very very huge schema.