topic Errors in vmstat output in Operating System - HP-UX

Errors in vmstat output

Carme Torca — Thu, 25 Mar 2010 10:18:12 GMT

Hi,

I have a system with 11.11, and when I do vmstat, the cpu shows numbers not correct like this:

/> vmstat 1 10
procs memory page faults cpu
r b w avm free re at pi po fr de sr in sy cs us sy id
2 0 0 181488 17002 73 42 1 0 0 0 0 1039 2639 135 38 7 55
2 0 0 181488 16961 8 2 0 0 0 0 0 110 528 61 16 10787569164366118900 88
2 0 0 181488 16961 7 1 0 0 0 0 0 111 524 62 19 12810238451404242900 83
2 0 0 181488 16961 77 11 0 0 0 0 0 108 1928 79 87 6 7
2 0 0 181488 16639 112 9 0 0 0 0 0 96 3844 81 22496029894529516000 9 22496029894529516000
2 0 0 181488 16977 89 7 0 0 0 0 0 99 3175 78 33539535679027937000 13 33539535679027937000
2 0 0 181488 16977 71 5 0 0 0 0 0 101 2572 74 65881228206384611000 25 65881228206384611000
2 0 0 181488 16977 56 4 0 0 0 0 0 105 2103 72 14299801829133778900 5 14299801829133778900
1 0 0 170364 16977 45 3 0 0 0 0 0 107 1784 70 18085043231078219800 7 119
1 0 0 170364 16977 36 2 0 0 0 0 0 111 1461 68 24595659497953821000 9 125

Could anyone help me? Maybe I need one patxe?

Thanks a lot of,
Carmen.

Re: Errors in vmstat output

Kanagaraj — Thu, 25 Mar 2010 10:31:31 GMT

vmstat output is not clear...take logs and attach that file.

Also use sar and top command.

If you have glance plus then use that for analysis.

Re: Errors in vmstat output

Carme Torca — Thu, 25 Mar 2010 11:00:21 GMT

Hi,

sar and top are correct, but not the vmstat.

Re: Errors in vmstat output

Turgay Cavdar — Thu, 25 Mar 2010 11:26:52 GMT

Hi,
Do you have the patch PHCO_25979 ?

http://www11.itrc.hp.com/service/patch/patchDetail.do?patchid=PHCO_25979&sel={hpux:11.11,}&BC=main|search|

( SR:8606229310 CR:JAGad98363 )
vmstat(1) is showing wrong values for percentage CPU times 'us', 'sy' and 'id' spent by all the active processors.The percentage cpu time spent in each state was calculated by storing the sum of cpu time spent in each state by all the active processors in a variable and dividing it by number of active processors. There was an overflow occurring while assigning the sum of cpu times spent by all the active processors. This problem can only occur on systems that are up for a long time.