Optimizing scheduling.

Jared Mauch (jared@wolverine.hq.cic.net)
Thu, 27 Jun 1996 16:15:26 -0400 (EDT)


I heard there are some optimizations in scheduling that I could
make to the kernel sources.. can someone point them to me?

My typical kernel state is about 30% system, which I find to be
really high. Here's a vmstat output, any help is greatly appreciated:

procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
8 2 0 16400 836 2284 7676 2 3 29 19 165 121 25 13 62
12 0 0 16400 772 1184 7236 11 0 336 3 628 1086 70 30 0
9 0 0 16400 772 940 6376 16 0 356 168 699 992 72 28 0
5 5 0 16388 788 892 6116 19 0 223 167 454 712 64 36 0
4 3 0 16388 768 852 5664 7 0 283 2 640 865 53 47 0
3 4 0 16512 752 852 5936 3 68 262 19 558 822 49 51 0
4 9 0 16764 1024 852 6128 4 68 82 171 401 681 56 44 0
6 6 0 18220 876 916 6412 43 488 172 199 634 969 55 45 0
9 0 0 18448 760 980 7284 35 80 442 67 965 1465 65 35 0
10 5 0 18328 768 1356 6864 103 0 385 170 1225 1611 62 38 0
10 7 0 17760 4092 1728 6904 203 0 310 542 1829 1841 62 38 0
11 4 0 16924 1944 1984 8452 108 0 483 48 961 1601 67 33 0
12 2 0 16832 820 1752 8860 16 0 457 11 794 1587 66 34 0
13 9 0 18088 768 1112 7268 164 447 1283 288 2213 3089 67 33 0
14 14 0 19720 744 736 6504 87 591 1900 489 2971 1936 62 38 0
16 6 0 19864 800 548 7268 76 271 1357 208 2083 2122 66 34 0
13 11 0 19852 776 592 6940 43 0 697 221 1505 1747 65 35 0
14 11 0 21452 972 648 6976 196 635 1380 271 2715 3324 62 38 0
5 28 1 22152 744 668 7704 45 235 443 323 638 1120 64 36 0
5 14 0 21884 1028 1388 7116 43 0 533 303 793 1454 61 39 0

- Jared