EXPERIMENT 1

This first experiment was run with the ìbusy waitî technique that we discussed (see CPU currentcy column).Ý Also, the resource containers received a huge influx of currentcy when the process began, causing disk spinup to occur immediately.Ý It was run three times, with three simultaneous processes (the same actual program) running at the same time.Ý With the idea of ensuring that disk spinup would occur by the third processí read request, they were given the respective priorities of 4, 5, and 6 on each run.Ý The read activity in each program was staggered ñ in the first process, it was activated immediately, the second process waited 5 sec, and the third process waited 10 sec.Ý The delay time between file reads in all three programs was 10 sec ñ not enough for the disk to spindown (total delay in each program was 30 sec).Ý The power allocation was 12000 mW.Ý Each program read in three files from disk, storing data on the time (including read time and 10 sec ìthink timesî) required to complete, the amount of currentcy used on the hard drive and the cpu, the total currentcy allocated to the resource container, and the residual currentcy in the container when the process ended.Ý This is where I am not totally sure if the statistics are accurate.Ý In situations where the HD currentcy spent is 0, it means the files were accidentally already in memory.

PID	Priority	Seconds	Microseconds	HD Currentcy	CPU Currentcy	Total allocation	Residual Currentcy
826	4	31	423935	2127444	15271457	14700000	7803854
855	4	31	339194	2267834	15533966	17300000	10001120
886	4	30	669354	2528378	14735931	14400000	7638446
827	5	30	319870	1622476	16005489	15100000	7974955
856	5	31	257410	1675641	16384136	17700000	10143308
887	5	30	349498	0	16803654	14400000	7596346
828	6	32	316225	1634887	17580285	15200000	6487748
857	6	31	359522	1584950	17367959	17800000	9350011
888	6	33	303738	2076900	17665827	14600000	5360358
	AVG	31	370971.7778	1724278.889	16372078.22	15688888.89	8039571.778
	MAX	33	669354	2528378	17665827	17800000	10143308
	MIN	30	257410	0	14735931	14400000	5360358

EXPERIMENT 2

In this second experiment, I eliminated the ìbusy waitî problem by using the sleep() system call.Ý This allowed more currentcy to become available for use in accessing the hard drive.Ý I also reset the amount of currentcy in the resource container to 0 after the process started, neutralizing the effects of the initial allocation of currentcy upon creation of the processí resource container.Ý This experiment was run 5 times, using respective priorities of 7, 8, and 9 for the three programs on each run.Ý The delay time between read requests was 60 seconds, allowing the hard drive to spindown after each set of requests (total delay of 180 sec for each program).Ý Since less currentcy was used by the CPU, the power allocation could be reduced to 3000 mW while still being sufficient to spinup the hard drive when two or three bids were added.Ý Once again, in situations where the amount of currentcy spent on the hard drive is 0, the files were accidentally stored in memory.Ý Oops!

PID	Priority	Seconds	Microseconds	HD Currentcy	CPU Currentcy	Total allocation	Residual Currentcy
837	7	192	425809	3110760	742867	21600000	16684534
924	7	181	432843	3323321	497494	20100000	15112385
951	7	180	152426	0	288081	19550000	18496933
982	7	180	170757	0	362912	20500000	18195792
1011	7	195	817041	3985701	1076602	21850000	15980855
839	8	189	724113	3219183	973809	20700000	16373775
926	8	189	333566	3742521	592202	21150000	16728595
952	8	190	63577	4540282	852344	21650000	16366854
983	8	188	293571	4414594	1020843	21550000	15905235
1012	8	187	559976	1053966	585539	21100000	18843660
838	9	183	484151	3158726	718952	21000000	16862910
925	9	182	793767	3696049	779209	21450000	16961720
953	9	183	623565	4464403	586952	21500000	16402913
984	9	181	959981	4406392	676489	21250000	16147328
1013	9	182	699979	4105396	521788	20650000	16264300
	AVG	185.4667	502341.4667	3148086.267	685072.2	21040000	16755185.93
	MAX	195	959981	4540282	1076602	21850000	18843660
	MIN	180	63577	0	288081	19550000	15112385