Problems with swapping

["Mindaugas Riauba" <mind bi lt>]

  Hello,

  System is fully updated RHES 3.0. And server seems to be
too aggressively swapping. "vmstat 1" output:

procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy id
wa
 0  1 612952  18932  11768 3377112   16    0   708   188  957  1037 14  4 73
8
 0  0 613684  20104  11768 3375008  108  752   788  1152 1058  1177 18  6 55
21
 1  0 613396  19176  11768 3375900  276    0   792   136  899   959  8  4 66
22
 1  3 613112  18272  11796 3376528  484    0   752   324  966  1124 12  6 66
15
 0  2 612744  16992  11800 3376688 1436   68  2220   188  762  1159  7  1 57
35
 0  2 613516  16920  11788 3372744 1028 1860  3116  2340  689  1183 13  3 41
43
 0  2 613560  20024  11796 3368632 1644  596  5268   700 1139  2119 21  4 35
39
 1  2 613408  17004  11792 3367168 1620    8  3772   144 1096  2085 18  5 36
41
 2  0 614288  16996  11816 3363604 1088 1168  2480  1772  864  1501 23  7 36
34
 0  1 615152  17084  11796 3362920 1004 2716  3692  3284  735  1410 25  4 38
33
 1  0 621212  16984  11760 3361676  464 6400  2592 20784  689   838 11  5 46
38
 0  1 624960  16984  11760 3358032  800  620  3908  6044  580  1034 16  5 39
40
 3  0 626268  16932  11780 3355664 1208  296  3652  4080  895  1703 16  5 30
49
 0  2 627104  16936  11812 3354812 1536 2068  4652  4836  749  1308 20  3 44
33
 1  1 628128  16952  11820 3356508  988 3312  2876  6708  621   992 12  4 45
40
 1  0 629280  16956  11828 3355056  744 1152  2324  3540  494   742 14  1 47
38
 1  0 629016  16992  11832 3355040  792  124  3380  1964  481   708 13  1 50
36

  Why to swap in/out when there is >3GB of data in cache? The single running
task
is Oracle. sysctl settings are default.

  What could I tune or where to read about reducing swapping?

  Thanks,

  Mindaugas