High bi/bo value of vmstat output

I have a system oracle 9204, ras 4.0, mem 4g, 4*cpu, online system

SGA_MAX_SIZE 3.2G
SHARED_POOL_SIZE 250M
DB_BUFFERS*DB_BLOCK_SIZE = 1.6g
PGA_AGGREAGE_SIZE=1.5g

I run vmstat to check the system, and find

[oracle_at_test04 iostat]$ vmstat 2 100
procs -----------memory---------- ---swap-- -----io---- --system--
----cpu----

 r b swpd free buff cache si so bi bo in   cs us sy id wa
 2 2 571320 7560 5088 2718912 0 0 42 1590 1200 449 0 1 87 12
 0 3 571528 3360 5116 2723772 114 452 156 6034 1215 746 3 1 82 14
 0 3 571392 96360 5124 2631792 126 320 186 8034 1202 1190 5 1 80 14
 0 1 571392 95020 5148 2633068 0 0 6 1468 1190 390 0 0 87 12
 0 1 571392 92996 5148 2634628 0 0 148 1452 1193 467 0 1 86 12
 0 1 571392 91588 5168 2636428 0 0 6 1496 1193 497 0 2 85 12
 1 1 570212 80580 5208 2647724 512 0 554 5924 1166 657 3  1 85 12
 0 1 569148 72964 5224 2655324 494 0 524 4136 1202 647 2  1 84 14
 0 1 569140 71816 5240 2656876 0 0 6 1724 1217 440 0 0 87 12
 1 3 568556 68488 5248 2659872 480 0 500 2336 1206 481 1 0 86 13
 0 3 568204 62768 5268 2664944 230 0 244 2826 1182 544 1 2 84 13
 0 2 567120 50736 5296 2676740 610 0 4222 2328 1176 458 2  0 85 12
 0 2 565716 42280 5304 2685772 602 0 758 4106 1186 601 2  0 84 13
 1 0 565592 38204 5328 2689948 42 0 44 2538 1210 471 0 0 87 12
 1 2 563572 28404 5348 2700120 984 0 1032 4466 1197 933 3 1 83 14

----io---- -

  bi bo
    0 3
   42 1590
  156 6034
  186 8034
    6 1468
  148 1452
    6 1496
  554 5924
  524 4136
    6 1724
  500 2336
  244 2826
 4222 2328
  758 4106
   44 2538
 1032 4466

The bi/bo has these values, is this bi/bo an issue? If yes, I need to reduce the sga size,
is that right?

Could you give suggestions?

Thanks Received on Fri Oct 06 2006 - 09:34:41 CDT