Here's a good question about tuning. Why does my system continuously scan (sr) to free up pages when the amount of "fre" pages is well ABOVE my minfree and maxfree settings:
$ vmo -a | grep free
maxfree = 2560
minfree = 2048
$
Here is sample output from a "vmstat -Itw 1" command:
kthr memory page faults cpu time
----------- --------------------- ------------------------------------ ------------------ ----------------------- --------
r b p avm fre fi fo pi po fr sr in sy cs us sy id wa pc ec hr mi se
5 0 0 1989487 5587 4443 1774 0 5 1492 3684 5980 78869 21788 73 12 13 2 2.80 93.4 10:16:23
5 1 0 1989487 5670 4217 1037 0 4 873 2002 5745 75376 19430 75 11 11 2 2.84 94.7 10:16:24
2 2 0 1989487 5640 4493 913 4 2 806 1132 5680 73061 18718 69 11 17 3 2.66 88.6 10:16:25
4 1 0 1989487 5665 4177 1428 1 7 1075 3309 6062 75888 20405 80 12 7 1 3.00 100.1 10:16:26
4 1 0 1989487 5658 4130 924 4 15 678 4199 5802 78884 19338 70 12 16 2 2.70 90.0 10:16:27
3 0 0 1989487 5639 4487 1002 0 5 777 3375 5877 79213 20233 78 12 9 1 2.93 97.8 10:16:28
2 2 0 1989487 5585 4493 1852 1 6 1419 3979 6225 77220 21150 76 13 9 2 2.90 96.8 10:16:29
4 1 0 1989487 5361 4444 1049 0 8 914 1291 6304 88256 21985 71 13 15 2 2.76 91.9 10:16:30
3 1 0 1989487 5633 4335 1038 1 8 782 1023 6040 80653 20273 80 12 7 1 2.98 99.5 10:16:31
I have been working with IBM and they suggested creating another paging space. That is just poopy talk because I don't want more paging space. I want it to quit scanning and work like it is suppose to! From what I understand it is NOT suppose to start scanning until the "fre" is less than the "minfree" and then it is suppose to scan and steal pages till the "fre" is equal to or greater than the "maxfree". Anyone have any ideas?