Yeah, I eventually want to add more realistic mathematical models to pull page refs from (exponential decline or something similar).
Using random page calls, the algorithms all perform approximately the same when there is a much higher discrepancy in the real world. When selecting algorithms to use in real use cases, the system page calls are monitored and then replayed into simulators like this one.
Still a few things to do before becoming practically usable, but it's good to compare the mechanisms of each paging algorithm and see how they work. I plan to add a configuration option so that it can be slowed to the point that someone could follow along with the algorithms.
Haha, yeah. It's from our second programming assignment where we were asked to implement 2 of the algorithms running one at a time, but I decided to completely rewrite it so I could compare all the algorithms side by side using varying methods.
4 comments
[ 0.20 ms ] story [ 17.5 ms ] threadUsing random page calls, the algorithms all perform approximately the same when there is a much higher discrepancy in the real world. When selecting algorithms to use in real use cases, the system page calls are monitored and then replayed into simulators like this one.
Still a few things to do before becoming practically usable, but it's good to compare the mechanisms of each paging algorithm and see how they work. I plan to add a configuration option so that it can be slowed to the point that someone could follow along with the algorithms.