There are certain details about data structures where we are making assumptions and it would be good if we could track how good the assumptions were.
For instance, the name server and clock server data structures. In the clock server, I have to make enough space for the doubly linked list to hold every possible task (well, I can shrink it a bit), but in practice we may never have more than 32 tasks that need/want to be delayed. Knowing when/if we hit this limit could speed up the clock server logic for managing the list.
To note, right now the stress test for the clock server does hit the clock server pretty hard in the sense that it has more than 32 things wanting delays.
There are certain details about data structures where we are making assumptions and it would be good if we could track how good the assumptions were.
For instance, the name server and clock server data structures. In the clock server, I have to make enough space for the doubly linked list to hold every possible task (well, I can shrink it a bit), but in practice we may never have more than 32 tasks that need/want to be delayed. Knowing when/if we hit this limit could speed up the clock server logic for managing the list.