and what about the origin behind the cdn? seems like that it’s still rather critical and would have migrated multiple physical hosts over the past 30 years.
This sort of thing is fascinating and I do something similar at a much smaller scale at day job, keeping a simple service running that enables the rest of the stack to survive.
No clue on QPS, but it can't be terribly high. It is a single request, compared to the 200+ requests served when someone loads something like CNN.com.
I doubt there is any origin at all. Since the response is so small and never changes it is likely hard coded into the config. Maybe it is a file on a shared object storage.
and what about the origin behind the cdn? seems like that it’s still rather critical and would have migrated multiple physical hosts over the past 30 years.
This sort of thing is fascinating and I do something similar at a much smaller scale at day job, keeping a simple service running that enables the rest of the stack to survive.