Высокий LA и низкая загруженность CPU

[Михаил Монашёв <postmaster@xxxxxxxxxxxxx>]

Здравствуйте nginx-ru,

Сложилась  непонятная  мне  ситуация.  Процессы похоже чего-то ожидают
постоянно.  Подскажите  пожалуйста,  чего  именно они ждут и как можно
исправить?

>uname -v
FreeBSD 6.2-STABLE

>vmstat 2
 procs      memory      page                    disks     faults      cpu
 r b w     avm    fre  flt  re  pi  po  fr  sr da0 da1   in   sy  cs us sy id
 0 12 0 3132612 235568 3533  17   0   0 4034 551   0   0 7681 3736 6746 36 10 54
 3 11 0 3197120 210964 6879   0   0   0 4287   0 183   5 30232 20268 7645 27 13 
59
 1 12 0 3185952 204024 2170   0   0   0 2526   0  34  25 31637 18288 6493 14  8 
78
 1 11 0 3172796 201860 3100   0   0   0 3840   0  62   7 31437 14755 6638 15  8 
77
 0 12 0 3138828 218780 3381   0   0   0 5715   0  34   9 31088 13534 5437 18  8 
74
 0 11 0 3140928 214688 2594   0   0   0 2378   0  76  29 31809 19866 6578 19  7 
74

>top
last pid: 13968;  load averages:  2.60,  2.12,  2.06  up 5+21:31:08  15:27:29
375 processes: 1 running, 374 sleeping
CPU states: 13.1% user,  0.4% nice,  4.8% system,  2.0% interrupt, 79.8% idle
Mem: 2183M Active, 1396M Inact, 256M Wired, 153M Cache, 112M Buf, 21M Free
Swap: 6144M Total, 6552K Used, 6137M Free

  PID THR PRI NICE   SIZE    RES STATE  C   TIME   WCPU COMMAND
13956   1   4    0 37276K 33188K accept 2   0:02  7.28% httpd
13958   1   4    0 22496K 21448K accept 0   0:01  5.52% httpd
25351  32  96    0   209M   152M ucond  0  74:10  3.71% mysqld
13862   1   4    0 15800K 14944K accept 2   0:04  3.04% httpd
13896   1   4    0 14624K 13768K accept 2   0:03  2.57% httpd
13849   1   4    0 37248K 33260K accept 0   0:02  0.93% httpd


>systat -vm
    2 users    Load  2.66  2.18  2.16                   3 фев 15:29

Mem:KB    REAL            VIRTUAL                       VN PAGER   SWAP PAGER
        Tot   Share      Tot    Share    Free           in   out     in   out
Act 2267216  178724  3182748   195584  190428  count
All 2420620  190476  5422480   213464          pages
Proc:                                                            Interrupts
  r   p   d   s   w   Csw  Trp  Sys  Int  Sof  Flt     70 cow    8097 total
  1       5 350       11k 6218  31k  63k 1385 4964   4852 zfod        em0 irq30
                                                     4811 ozfod    97 ahd1 irq49
 6.3%Sys   1.5%Intr 15.8%User  0.0%Nice 76.4%Idle      99%ozfod       ahd0 irq50
|    |    |    |    |    |    |    |    |    |    |       daefr  4000 cpu0: time
===+>>>>>>>>                                            6 prcfr       cpu3: time
                                       465 dtbuf     4616 totfr       cpu1: time
Namei     Name-cache   Dir-cache    100000 desvn          react  4000 cpu2: time
   Calls    hits   %    hits   %     63092 numvn          pdwak
   12007   11941  99                 24833 frevn          pdpgs
                                                          intrn
Disks   da0   da1   da2                            263808 wire
KB/t  14.93 50.25 55.04                           2252644 act
tps      27    11     5                           1398252 inact
MB/s   0.40  0.56  0.27                            145564 cache
 %busy    11    11     4                             44864 free
                                                   114784 buf

>sysctl -a | grep kern.polling
kern.polling.idlepoll_sleeping: 1
kern.polling.stalled: 4
kern.polling.suspect: 839617
kern.polling.phase: 0
kern.polling.enable: 0
kern.polling.handlers: 1
kern.polling.residual_burst: 0
kern.polling.pending_polls: 0
kern.polling.lost_polls: 1160401
kern.polling.short_ticks: 2457
kern.polling.reg_frac: 20
kern.polling.user_frac: 50
kern.polling.idle_poll: 0
kern.polling.each_burst: 5
kern.polling.burst_max: 150
kern.polling.burst: 145

>netstat -m
1957/1013/2970 mbufs in use (current/cache/total)
790/372/1162/33792 mbuf clusters in use (current/cache/total/max)
790/304 mbuf+clusters out of packet secondary zone in use (current/cache)
0/0/0/0 4k (page size) jumbo clusters in use (current/cache/total/max)
0/0/0/0 9k jumbo clusters in use (current/cache/total/max)
0/0/0/0 16k jumbo clusters in use (current/cache/total/max)
2069K/997K/3066K bytes allocated to network (current/cache/total)
0/0/0 requests for mbufs denied (mbufs/clusters/mbuf+clusters)
0/0/0 requests for jumbo clusters denied (4k/9k/16k)
783/1662/8704 sfbufs in use (current/peak/max)
0 requests for sfbufs denied
0 requests for sfbufs delayed
2318691 requests for I/O initiated by sendfile
11993 calls to protocol drain routines

>iostat 2
      tty             da0              da1              da2             cpu
 tin tout  KB/t tps  MB/s   KB/t tps  MB/s   KB/t tps  MB/s  us ni sy in id
   0  143 18.36  42  0.74  46.19  23  1.03  49.24  13  0.65  22 13  9  2 54
   0  207 20.15  26  0.51   0.00   0  0.00  36.17  11  0.41  17  1  9  0 73
   0  496 19.40  42  0.80  65.78   4  0.29  25.59  48  1.21  21  1 10  1 68
   0  404 16.57  54  0.88  50.00   4  0.20  42.18   5  0.23  20  0  6  1 73
   0  313  9.56  41  0.38   0.00   0  0.00  49.23   6  0.31  23  0  6  1 70
   0   38 15.22  33  0.50  45.89  38  1.70  24.00   2  0.05  17  0  9  0 73
   0  404 21.76  41  0.88  51.76  25  1.29  58.67   6  0.34  12  0  8  1 79

>netstat -Lan | grep 80
tcp4  0/0/128        127.0.1.18.80
tcp4  10/0/128       127.0.1.17.80
tcp4  0/0/128        127.0.1.16.80
tcp4  0/0/128        127.0.1.15.80
tcp4  0/0/128        127.0.1.14.80
tcp4  0/0/128        127.0.1.13.80
tcp4  0/0/128        127.0.1.12.80
tcp4  0/0/128        127.0.1.10.80
tcp4  0/0/128        127.0.1.11.80
tcp4  0/0/128        127.0.1.8.80
tcp4  0/0/128        127.0.1.5.80
tcp4  0/1/128        127.0.1.6.80
tcp4  0/0/128        127.0.1.4.80
tcp4  0/0/128        127.0.1.2.80
tcp4  0/0/128        217.16.26.114.80
tcp4  0/0/128        217.16.27.103.80
tcp4  0/0/128        217.16.27.104.80


С уважением,
Михаил Монашёв, SoftSearch.ru
Member of Independent Software Developers Forum (ISDEF)
mailto:postmaster@xxxxxxxxxxxxx
ICQ# 166233339
http://softsearch.ru/
Без бэкапа по жизни.