[FEATURE]: reading trace based on clock_time of each request

[ddkkpp]

It seems libCacheSim does not support reading trace based on clock_time of each request. The current code logic is reading the next request as soon as the current request has been processed. In this case, a simulation for any algorithm occupies 100% CPU at one core. Consequently, we cannot distinguish the CPU utilization of different algorithms. Do you think it is necessary to read trace based on clock_time or is there another way to measure the CPU utilizatio?

[1a1a11a]

Can you describe a use case of this new emulator? Is the goal to measure CPU utilization of an eviction algorithm? Or do you want to emulate a full storage system?

[ddkkpp]

Yeah， the goal is to measure CPU utilization of an eviction algorithm.

[1a1a11a]

The issue is that trace reading will dominate the CPU utilization.