system utilities job

[Коньков Евгений]

>> From owner-freebsd-questions at freebsd.org  Sat Nov 13 13:49:32 2010
>> Date: Sat, 13 Nov 2010 19:13:38 +0200
>> From: =?windows-1251?B?yu7t/Oru4iDF4uPl7ejp?= <kes-kes at yandex.ru>
>> To: freebsd-questions at freebsd.org
>> Subject: system utilities job
>>
>> Hi, Freebsd-questions.
>>
>> why I have different results for re0 interface for interrupt
>> statistic?
>>
>> # vmstat -i
>> interrupt                          total       rate
>> irq1: atkbd0                           5          0
>> irq16: ehci0                       84414          1
>> irq17: atapci0                    329838          5
>> irq23: ehci1                      112524          2
>> cpu0:timer                     119506195       2125
>> irq266: re0                    492258114       8753   <<<<<<<<<<
>> cpu3:timer                      98445240       1750
>> cpu1:timer                      95782407       1703
>> cpu2:timer                      89616119       1593
>> Total                          896134856      15934
>>
>> #systat -v
>>     3 users    Load  0,42  0,51  0,51                  13 ноя 19:12
>>
>> Mem:KB    REAL            VIRTUAL                       VN PAGER   SWAP PAGER
>>         Tot   Share      Tot    Share    Free           in   out     in   out
>> Act   86084    4912   446820     6376 1474852  count
>> All  207628    8180  2620768    32204          pages
>> Proc:                                                            Interrupts
>>   r   p   d   s   w   Csw  Trp  Sys  Int  Sof  Flt        cow   21237 total
>>              40       93k   12 1129  12k  40k   15      1 zfod        atkbd0 1
>>                                                           ozfod     1 ehci0 16
>>  2,8%Sys  10,9%Intr  0,0%User  0,0%Nice 86,2%Idle        %ozfod    10 atapci0 17
>> |    |    |    |    |    |    |    |    |    |    |       daefr     2 ehci1 23
>> =++++++                                                   prcfr  2126 cpu0:timer
>>                                         30 dtbuf        3 totfr 12744 re0 266         <<<<<<<<<<
>> Namei     Name-cache   Dir-cache    110674 desvn          react  2127 cpu3:timer
>>    Calls    hits   %    hits   %     71986 numvn          pdwak  2123 cpu1:timer
>>        7       7 100                 27160 frevn          pdpgs  2104 cpu2:timer
>>                                                           intrn
>> Disks   ad4                                        172156 wire
>> KB/t  15,84                                         49524 act
>> tps      10                                        272900 inact
>> MB/s   0,16                                           272 cache
>> %busy     0                                       1474580 free
>>
>>

RB> for the same reason _all_ the interrupt counts are different.  You ran two
RB> different commands. at (slightly) differnt times.  Thus, the 'real-time'
RB> events being reported are _differnet_.   A different amount of network
RB> traffic will result in different numbers of interrupts from the ethernet
RB> interface.   Entirely expected.

I have ran both commands in two different shell simulatenously at same time.
This OS is 9CURRENT. In 7-2RELEASE interupts was same.




-- 
С уважением,
 Коньков                          mailto:kes-kes at yandex.ru