I don't know what to call this section

Let T_r be the minimum response time for an operation of any sort, T_f be the time a handle must sit before being moved back to the free list, and N_tot be the total number of handles available on a server.

The pathological case would be one where a caller

• fills up the recently_freed list
• immediately starts consuming handles as quickly as possible to make for the largest possible recently_freed list in the next pass

This results in the largest number of handles being unavailable due to sitting on the overflow_list. Call N_purg the number of handles waiting in ``purgatory'' ( waiting for T_f to pass)

We should try to collect statistics and see what T_r and N_purg end up being for real and pathological workloads.