Disk Scheduler for the 2.6 kernel

Few days ago I started on an ambitious project of implementing a new disk scheduler for the 2.6 Linux kernel. The scheduler was to make use of the T-Algorithm . This involved some learning curve and I still haven't done much on it. Also with Bonda being at my place I could hardly get to it.

I have also being pondering about how much of practical use T-Algorithm is, since it is of order O(n^2). I am planning to instead implement the CSCAN disk scheduling algorithm which by making use of the Red Black Trees will have the time complexity of O(log n). However my earlier experiments did show that the average wait time in case of T-Algorithm is better than the CSCAN algorithm