hi all
I got a question about avmin vmstat. the output will look like the
following:
kthr memory page faults cpu

-------- ----------- ------------------------ ------------
-----------
r b p avm fre fi fo pi po fr sr in sy cs us sy
id wa



6 1 0 1204592 226324 0 27 0 0 0 0 4925 29964 5809 11 6
74 9 07:55
3 1 1 1203968 226470 0 38 0 0 0 0 5957 38662 7639 18 8
51 22 08:00
4 1 1 1205824 224274 0 28 0 0 0 0 5013 33039 6274 15 6
60 18 08:05
6 1 0 1212255 184874 0 147 0 0 0 0 7158 44726 8482 20 9
56 15 08:10
10 1 0 1213204 180195 11 47 0 0 0 0 6450 104687 7426 22 18
49 11 08:15
3 1 0 1218707 174248 0 39 0 0 0 0 6507 40434 7169 21 8
54 18 08:20
6 1 0 1218778 173697 0 45 0 0 0 0 6537 42181 7507 20 9
58 13 08:25
6 1 0 1223190 168833 0 38 0 0 0 0 6241 38738 7058 18 8
56 18 08:30
3 1 1 1224145 134853 0 147 0 0 0 0 6954 41469 7975 22 9
49 21 08:35
5 1 0 1224091 134573 0 34 0 0 0 0 5956 39887 8113 19 8
57 16 08:40
6 1 1 1224757 133558 0 50 0 0 0 0 8919 55389 10889 27 12
45 17 08:45
7 1 1 1231535 125674 2 37 0 0 0 0 6590 41358 8161 21 9
50 21 08:50
6 1 1 1238121 86059 0 152 0 0 0 0 7102 44994 7911 21 10
51 19 08:55
10 1 1 1235240 88330 0 57 0 0 0 0 8413 58101 11648 27 12
38 23 09:00
5 1 1 1235376 87702 0 47 0 0 0 0 7741 49890 9447 26 11
38 25 09:05
8 1 0 1251154 71593 0 35 0 0 0 0 6391 39332 7065 18 8
56 17 09:10
7 1 0 1252791 36988 0 153 0 0 0 0 6755 41707 7541 24 8
55 13 09:15
9 1 1 1251838 37641 0 45 0 0 0 0 7596 47724 9433 25 11
46 18 09:20
11 1 0 1251056 38057 0 38 0 0 0 0 6787 44773 8994 19 10
57 15 09:25
5 1 2 1249743 39040 0 42 0 0 0 0 7241 43201 7893 19 10
41 30 09:30
3 1 4 1251357 36995 0 43 0 0 0 0 6819 40924 8272 20 9
23 48 09:35
5 1 3 1253612 1802 0 145 0 0 0 0 5937 38136 8122 17 9
31 43 09:40
4 1 2 1253914 1332 0 32 0 0 0 0 5572 34705 6595 16 7
36 41 09:45
6 1 2 1254596 2099 0 35 0 0 5 9 6499 42685 8585 20 9
37 34 09:50
7 1 1 1258113 768 1 50 0 0 9 20 6722 46685 9081 23 10
38 28 09:55


the tunning book said if during this busy period the number of active
vitual pages (avm) are consistently less than the number of pages of
real memory, then maxperm should be lowered to shrink the size of the
system's fielsystem cache.

but actually the maxperm and minperm in my server are 30% and 10%, it
should be ok for my environment, but even I lowered the number of
maxperm, avm still keep low.

the output of vmtune showed as below:

vmtune
vmtune: current values:
-p -P -r -R -f -F -N
-W
minperm maxperm minpgahead maxpgahead minfree maxfree pd_npages
maxrandwrt
200539 601617 2 8 120 128 65536
0

-M -w -k -c -b -B -u
-l -d
maxpin npswarn npskill numclust numfsbufs hd_pbuf_cnt lvm_bufcnt
lrubucket defps
1677709 83456 20864 1 186 896 30
131072 1

-s -n -S -L -g
-h
sync_release_ilock nokilluid v_pinshm lgpg_regions lgpg_size
strict_maxperm
0 0 0 0 0 0

-t -j -J -z
maxclient j2_nPagesPer j2_maxRandomWrite j2_nRandomCluster
401078 32 0 0 1611497499

-Z -q -Q -y
j2_nBufferPer j2_minPageReadAhead j2_maxPageReadAhead
memory_affinity
512 2 8 0

-V -i -e -E
num_spec_dataseg spec_dataseg_int jfs_clread_enabled
jfs_use_read_lock
0 512 0 1

PTA balance threshold percentage = 50.0%

number of valid memory pages = 2097136 maxperm=30.0% of real memory
maximum pinable=80.0% of real memory minperm=10.0% of real memory
number of file memory pages = 1081882 numperm=53.9% of real memory
number of compressed memory pages = 0 compressed=0.0% of real memory
number of client memory pages = 0 numclient=0.0% of real memory
# of remote pgs sched-pageout = 0 maxclient=20.0% of real memory


any idea?

thanks in advance!

Frank

another thing is the "fo" in output of vmstat.
the colum of "fo" constantly has nonzero number in all time. does it
mean anything?
I could not find any helpful information about the "fo" colume in IBM's
website and the only thing is the concept of "fo".

please give me a light!


thanks

Frank