Cargas altas do Linux com baixo uso de CPU / memória

12

Eu tenho uma situação bastante estranha, onde minhas cargas de caixa do CentOS 5.5 são altas, mas a CPU e a memória usadas são muito baixas:

top - 20:41:38 up 42 days,  6:14,  2 users,  load average: 19.79, 21.25, 18.87
Tasks: 254 total,   1 running, 253 sleeping,   0 stopped,   0 zombie
Cpu(s):  3.8%us,  0.3%sy,  0.1%ni, 95.0%id,  0.6%wa,  0.0%hi,  0.1%si,  0.0%st
Mem:   4035284k total,  4008084k used,    27200k free,    38748k buffers
Swap:  4208928k total,   242576k used,  3966352k free,  1465008k cached

free -mt
             total       used       free     shared    buffers     cached
Mem:          3940       3910         29          0         37       1427
-/+ buffers/cache:       2445       1495
Swap:         4110        236       3873
Total:        8050       4147       3903

O Iostat também mostra bons resultados:

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
           3.83    0.13    0.41    0.58    0.00   95.05

Aqui está a saída ps aux:

USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
root         1  0.0  0.0  10348    80 ?        Ss    2010   2:11 init [3]                                           
root         2  0.0  0.0      0     0 ?        S<    2010   0:00 [migration/0]
root         3  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/0]
root         4  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/0]
root         5  0.0  0.0      0     0 ?        S<    2010   0:02 [migration/1]
root         6  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/1]
root         7  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/1]
root         8  0.0  0.0      0     0 ?        S<    2010   0:02 [migration/2]
root         9  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/2]
root        10  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/2]
root        11  0.0  0.0      0     0 ?        S<    2010   0:02 [migration/3]
root        12  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/3]
root        13  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/3]
root        14  0.0  0.0      0     0 ?        S<    2010   0:03 [migration/4]
root        15  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/4]
root        16  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/4]
root        17  0.0  0.0      0     0 ?        S<    2010   0:01 [migration/5]
root        18  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/5]
root        19  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/5]
root        20  0.0  0.0      0     0 ?        S<    2010   0:11 [migration/6]
root        21  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/6]
root        22  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/6]
root        23  0.0  0.0      0     0 ?        S<    2010   0:01 [migration/7]
root        24  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/7]
root        25  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/7]
root        26  0.0  0.0      0     0 ?        S<    2010   0:00 [migration/8]
root        27  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/8]
root        28  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/8]
root        29  0.0  0.0      0     0 ?        S<    2010   0:00 [migration/9]
root        30  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/9]
root        31  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/9]
root        32  0.0  0.0      0     0 ?        S<    2010   0:08 [migration/10]
root        33  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/10]
root        34  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/10]
root        35  0.0  0.0      0     0 ?        S<    2010   0:05 [migration/11]
root        36  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/11]
root        37  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/11]
root        38  0.0  0.0      0     0 ?        S<    2010   0:02 [migration/12]
root        39  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/12]
root        40  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/12]
root        41  0.0  0.0      0     0 ?        S<    2010   0:14 [migration/13]
root        42  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/13]
root        43  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/13]
root        44  0.0  0.0      0     0 ?        S<    2010   0:04 [migration/14]
root        45  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/14]
root        46  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/14]
root        47  0.0  0.0      0     0 ?        S<    2010   0:01 [migration/15]
root        48  0.0  0.0      0     0 ?        SN    2010   0:00 [ksoftirqd/15]
root        49  0.0  0.0      0     0 ?        S<    2010   0:00 [watchdog/15]
root        50  0.0  0.0      0     0 ?        S<    2010   0:00 [events/0]
root        51  0.0  0.0      0     0 ?        S<    2010   0:00 [events/1]
root        52  0.0  0.0      0     0 ?        S<    2010   0:00 [events/2]
root        53  0.0  0.0      0     0 ?        S<    2010   0:00 [events/3]
root        54  0.0  0.0      0     0 ?        S<    2010   0:00 [events/4]
root        55  0.0  0.0      0     0 ?        S<    2010   0:00 [events/5]
root        56  0.0  0.0      0     0 ?        S<    2010   0:00 [events/6]
root        57  0.0  0.0      0     0 ?        S<    2010   0:00 [events/7]
root        58  0.0  0.0      0     0 ?        S<    2010   0:00 [events/8]
root        59  0.0  0.0      0     0 ?        S<    2010   0:00 [events/9]
root        60  0.0  0.0      0     0 ?        S<    2010   0:00 [events/10]
root        61  0.0  0.0      0     0 ?        S<    2010   0:00 [events/11]
root        62  0.0  0.0      0     0 ?        S<    2010   0:00 [events/12]
root        63  0.0  0.0      0     0 ?        S<    2010   0:00 [events/13]
root        64  0.0  0.0      0     0 ?        S<    2010   0:00 [events/14]
root        65  0.0  0.0      0     0 ?        S<    2010   0:00 [events/15]
root        66  0.0  0.0      0     0 ?        S<    2010   0:00 [khelper]
root       107  0.0  0.0      0     0 ?        S<    2010   0:00 [kthread]
root       126  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/0]
root       127  0.0  0.0      0     0 ?        S<    2010   0:03 [kblockd/1]
root       128  0.0  0.0      0     0 ?        S<    2010   0:01 [kblockd/2]
root       129  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/3]
root       130  0.0  0.0      0     0 ?        S<    2010   0:05 [kblockd/4]
root       131  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/5]
root       132  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/6]
root       133  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/7]
root       134  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/8]
root       135  0.0  0.0      0     0 ?        S<    2010   0:02 [kblockd/9]
root       136  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/10]
root       137  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/11]
root       138  0.0  0.0      0     0 ?        S<    2010   0:04 [kblockd/12]
root       139  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/13]
root       140  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/14]
root       141  0.0  0.0      0     0 ?        S<    2010   0:00 [kblockd/15]
root       142  0.0  0.0      0     0 ?        S<    2010   0:00 [kacpid]
root       281  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/0]
root       282  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/1]
root       283  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/2]
root       284  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/3]
root       285  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/4]
root       286  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/5]
root       287  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/6]
root       288  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/7]
root       289  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/8]
root       290  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/9]
root       291  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/10]
root       292  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/11]
root       293  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/12]
root       294  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/13]
root       295  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/14]
root       296  0.0  0.0      0     0 ?        S<    2010   0:00 [cqueue/15]
root       299  0.0  0.0      0     0 ?        S<    2010   0:00 [khubd]
root       301  0.0  0.0      0     0 ?        S<    2010   0:00 [kseriod]
root       490  0.0  0.0      0     0 ?        S     2010   0:00 [khungtaskd]
root       493  0.1  0.0      0     0 ?        S<    2010  94:48 [kswapd1]
root       494  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/0]
root       495  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/1]
root       496  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/2]
root       497  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/3]
root       498  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/4]
root       499  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/5]
root       500  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/6]
root       501  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/7]
root       502  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/8]
root       503  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/9]
root       504  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/10]
root       505  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/11]
root       506  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/12]
root       507  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/13]
root       508  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/14]
root       509  0.0  0.0      0     0 ?        S<    2010   0:00 [aio/15]
root       665  0.0  0.0      0     0 ?        S<    2010   0:00 [kpsmoused]
root       808  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/0]
root       809  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/1]
root       810  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/2]
root       811  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/3]
root       812  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/4]
root       813  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/5]
root       814  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/6]
root       815  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/7]
root       816  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/8]
root       817  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/9]
root       818  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/10]
root       819  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/11]
root       820  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/12]
root       821  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/13]
root       822  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/14]
root       823  0.0  0.0      0     0 ?        S<    2010   0:00 [ata/15]
root       824  0.0  0.0      0     0 ?        S<    2010   0:00 [ata_aux]
root       842  0.0  0.0      0     0 ?        S<    2010   0:00 [scsi_eh_0]
root       843  0.0  0.0      0     0 ?        S<    2010   0:00 [scsi_eh_1]
root       844  0.0  0.0      0     0 ?        S<    2010   0:00 [scsi_eh_2]
root       845  0.0  0.0      0     0 ?        S<    2010   0:00 [scsi_eh_3]
root       846  0.0  0.0      0     0 ?        S<    2010   0:00 [scsi_eh_4]
root       847  0.0  0.0      0     0 ?        S<    2010   0:00 [scsi_eh_5]
root       882  0.0  0.0      0     0 ?        S<    2010   0:00 [kstriped]
root       951  0.0  0.0      0     0 ?        S<    2010   4:24 [kjournald]
root       976  0.0  0.0      0     0 ?        S<    2010   0:00 [kauditd]
postfix    990  0.0  0.0  54208  2284 ?        S    21:19   0:00 pickup -l -t fifo -u
root      1013  0.0  0.0  12676     8 ?        S<s   2010   0:00 /sbin/udevd -d
root      1326  0.0  0.0  90900  3400 ?        Ss   14:53   0:00 sshd: root@notty 
root      1410  0.0  0.0  53972  2108 ?        Ss   14:53   0:00 /usr/libexec/openssh/sftp-server
root      2690  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/0]
root      2691  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/1]
root      2692  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/2]
root      2693  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/3]
root      2694  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/4]
root      2695  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/5]
root      2696  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/6]
root      2697  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/7]
root      2698  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/8]
root      2699  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/9]
root      2700  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/10]
root      2701  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/11]
root      2702  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/12]
root      2703  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/13]
root      2704  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/14]
root      2705  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpathd/15]
root      2706  0.0  0.0      0     0 ?        S<    2010   0:00 [kmpath_handlerd]
root      2755  0.0  0.0      0     0 ?        S<    2010   4:35 [kjournald]
root      2757  0.0  0.0      0     0 ?        S<    2010   3:38 [kjournald]
root      2759  0.0  0.0      0     0 ?        S<    2010   4:10 [kjournald]
root      2761  0.0  0.0      0     0 ?        S<    2010   4:26 [kjournald]
root      2763  0.0  0.0      0     0 ?        S<    2010   3:15 [kjournald]
root      2765  0.0  0.0      0     0 ?        S<    2010   3:04 [kjournald]
root      2767  0.0  0.0      0     0 ?        S<    2010   3:02 [kjournald]
root      2769  0.0  0.0      0     0 ?        S<    2010   2:58 [kjournald]
root      2771  0.0  0.0      0     0 ?        S<    2010   0:00 [kjournald]
root      3340  0.0  0.0   5908   356 ?        Ss    2010   2:48 syslogd -m 0
root      3343  0.0  0.0   3804   212 ?        Ss    2010   0:03 klogd -x
root      3430  0.0  0.0      0     0 ?        S<    2010   0:50 [kondemand/0]
root      3431  0.0  0.0      0     0 ?        S<    2010   0:54 [kondemand/1]
root      3432  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/2]
root      3433  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/3]
root      3434  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/4]
root      3435  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/5]
root      3436  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/6]
root      3437  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/7]
root      3438  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/8]
root      3439  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/9]
root      3440  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/10]
root      3441  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/11]
root      3442  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/12]
root      3443  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/13]
root      3444  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/14]
root      3445  0.0  0.0      0     0 ?        S<    2010   0:00 [kondemand/15]
root      3461  0.0  0.0  10760   284 ?        Ss    2010   3:44 irqbalance
rpc       3481  0.0  0.0   8052     4 ?        Ss    2010   0:00 portmap
root      3526  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/0]
root      3527  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/1]
root      3528  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/2]
root      3529  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/3]
root      3530  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/4]
root      3531  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/5]
root      3532  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/6]
root      3533  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/7]
root      3534  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/8]
root      3535  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/9]
root      3536  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/10]
root      3537  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/11]
root      3538  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/12]
root      3539  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/13]
root      3540  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/14]
root      3541  0.0  0.0      0     0 ?        S<    2010   0:00 [rpciod/15]
root      3563  0.0  0.0  10160     8 ?        Ss    2010   0:00 rpc.statd
root      3595  0.0  0.0  55180     4 ?        Ss    2010   0:00 rpc.idmapd
dbus      3618  0.0  0.0  21256    28 ?        Ss    2010   0:00 dbus-daemon --system
root      3649  0.2  0.4 563084 18796 ?        S<sl  2010 179:03 mfsmount /mnt/mfs -o rw,mfsmaster=web1.ovs.local
root      3702  0.0  0.0   3800     8 ?        Ss    2010   0:00 /usr/sbin/acpid
68        3715  0.0  0.0  31312   816 ?        Ss    2010   3:14 hald
root      3716  0.0  0.0  21692    28 ?        S     2010   0:00 hald-runner
68        3726  0.0  0.0  12324     8 ?        S     2010   0:00 hald-addon-acpi: listening on acpid socket /var/run/acpid.socket
68        3730  0.0  0.0  12324     8 ?        S     2010   0:00 hald-addon-keyboard: listening on /dev/input/event0
root      3773  0.0  0.0  62608   332 ?        Ss    2010   0:00 /usr/sbin/sshd
ganglia   3786  0.0  0.0  24704   988 ?        Ss    2010  14:26 /usr/sbin/gmond
root      3843  0.0  0.0  54144   300 ?        Ss    2010   1:49 /usr/libexec/postfix/master
postfix   3855  0.0  0.0  54860  1060 ?        S     2010   0:22 qmgr -l -t fifo -u
root      3877  0.0  0.0  74828   708 ?        Ss    2010   1:15 crond
root      3891  1.4  1.9 326960 77704 ?        S<l   2010 896:59 mfschunkserver
root      4122  0.0  0.0  18732   176 ?        Ss    2010   0:10 /usr/sbin/atd
root      4193  0.0  0.8 129180 35984 ?        Ssl   2010  11:04 /usr/bin/ruby /usr/sbin/puppetd
root      4223  0.0  0.0  18416   172 ?        S     2010   0:10 /usr/sbin/smartd -q never
root      4227  0.0  0.0   3792     8 tty1     Ss+   2010   0:00 /sbin/mingetty tty1
root      4230  0.0  0.0   3792     8 tty2     Ss+   2010   0:00 /sbin/mingetty tty2
root      4231  0.0  0.0   3792     8 tty3     Ss+   2010   0:00 /sbin/mingetty tty3
root      4233  0.0  0.0   3792     8 tty4     Ss+   2010   0:00 /sbin/mingetty tty4
root      4234  0.0  0.0   3792     8 tty5     Ss+   2010   0:00 /sbin/mingetty tty5
root      4236  0.0  0.0   3792     8 tty6     Ss+   2010   0:00 /sbin/mingetty tty6
root      5596  0.0  0.0  19368    20 ?        Ss    2010   0:00 DarwinStreamingServer
qtss      5597  0.8  0.9 166572 37408 ?        Sl    2010 523:02 DarwinStreamingServer
root      8714  0.0  0.0      0     0 ?        S    Jan31   0:33 [pdflush]
root      9914  0.0  0.0  65612   968 pts/1    R+   21:49   0:00 ps aux
root     10765  0.0  0.0  76792  1080 ?        Ss   Jan24   0:58 SCREEN
root     10766  0.0  0.0  66212   872 pts/3    Ss   Jan24   0:00 /bin/bash
root     11833  0.0  0.0  63852  1060 pts/3    S+   17:17   0:00 /bin/sh ./launch.sh
root     11834  437 42.9 4126884 1733348 pts/3 Sl+  17:17 1190:50 /usr/bin/java -Xms128m -Xmx512m -XX:+UseConcMarkSweepGC -jar /JavaCore/JavaCore.jar
root     13127  4.7  1.1 110564 46876 ?        Ssl  17:18  12:55 /JavaCore/fetcher.bin
root     19392  0.0  0.0  90108  3336 ?        Rs   20:35   0:00 sshd: root@pts/1 
root     19401  0.0  0.0  66216  1640 pts/1    Ss   20:35   0:00 -bash
root     20567  0.0  0.0  90108   412 ?        Ss   Jan16   1:58 sshd: root@pts/0 
root     20569  0.0  0.0  66084   912 pts/0    Ss   Jan16   0:00 -bash
root     21053  0.0  0.0  63856    28 ?        S    Jan30   0:00 /bin/sh /usr/bin/WowzaMediaServerd /usr/local/WowzaMediaServer/bin/setenv.sh /var/run/WowzaM
root     21054  2.9 10.3 2252652 418468 ?      Sl   Jan30 314:25 java -Xmx1200M -server -Djava.net.preferIPv4Stack=true -Dcom.sun.management.jmxremote=true -
root     21915  0.0  0.0      0     0 ?        S    Feb01   0:00 [pdflush]
root     29996  0.0  0.0  76524  1004 pts/0    S+   14:41   0:00 screen -x

saída sar -W:

12:00:01 AM  pswpin/s pswpout/s
12:10:01 AM      0.00      0.00
12:20:01 AM      0.00      0.00
12:30:02 AM      0.00      0.00
12:40:01 AM      0.00      0.00
12:50:01 AM      0.00      0.00
01:00:01 AM      0.00      0.00
01:10:01 AM      0.00      0.00
01:20:01 AM      0.00      0.00
01:30:01 AM      0.00      0.00
01:40:01 AM      0.00      0.00
01:50:01 AM      0.00      0.00
02:00:02 AM      0.00      0.00
02:10:01 AM      0.07      0.00
02:20:01 AM      0.00      0.00
02:30:02 AM      0.00      0.00
02:40:01 AM      0.00      0.00
02:50:01 AM      0.00      0.00
03:00:01 AM      0.00      0.00
03:10:01 AM      0.00      0.00
03:20:01 AM      0.00      0.00
03:30:01 AM      0.00      0.00
03:40:02 AM      0.00      0.00
03:50:01 AM      0.00      0.00
04:00:01 AM      0.00      0.00
04:10:01 AM      0.00      0.00
04:20:02 AM      0.01      0.00
04:30:01 AM      0.00      0.00
04:40:02 AM      0.11      0.00
04:50:01 AM      0.01      0.00
05:00:02 AM      0.03      0.00
05:10:01 AM      0.00      0.00
05:20:02 AM      0.01      0.00
05:30:01 AM      0.04      0.00
05:40:02 AM      0.08      0.00
05:50:01 AM      0.00      0.00
06:00:02 AM      0.11      0.00
06:10:01 AM      0.01      0.00
06:20:01 AM      0.00      0.00
06:30:02 AM      0.00      0.00
06:40:01 AM      0.05      0.00
06:50:02 AM      0.00      0.00
07:00:02 AM      0.01      0.00
07:10:01 AM      0.02      0.00
07:20:02 AM      0.00      0.00
07:30:01 AM      0.00      0.00
07:40:02 AM      0.17      0.00
07:50:01 AM      0.11      0.00
08:00:02 AM      0.00      0.00
08:10:01 AM      0.04      0.00
08:20:02 AM      0.00      0.00
08:30:01 AM      0.00      0.00
08:40:02 AM      0.03      0.00
08:50:01 AM      0.00      0.00
09:00:02 AM      0.08      0.00
09:10:01 AM      0.00      0.00
Average:         0.02      0.00

saída sar -d 5 0

09:18:40 AM       DEV       tps  rd_sec/s  wr_sec/s  avgrq-sz  avgqu-sz     await     svctm     %util
09:18:45 AM    dev8-0     21.96     11.18   2128.54     97.45      0.62     20.33      1.75      3.85
09:18:45 AM    dev8-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM    dev8-2     18.16      0.00    507.78     27.96      0.39     21.42      1.09      1.98
09:18:45 AM    dev8-3      3.79     11.18   1620.76    430.32      0.23     15.11      4.95      1.88
09:18:45 AM    dev8-4      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM    dev8-5      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM   dev8-16     19.36      0.00    619.56     32.00      0.35     18.22      0.94      1.82
09:18:45 AM   dev8-17     17.96      0.00    594.01     33.07      0.35     19.61      0.99      1.78
09:18:45 AM   dev8-18      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM   dev8-19      1.40      0.00     25.55     18.29      0.00      0.29      0.29      0.04
09:18:45 AM   dev8-20      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM   dev8-21      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM   dev8-32     25.55    742.51    846.31     62.19      0.69     27.04      1.63      4.17
09:18:45 AM   dev8-33     22.55      0.00    844.71     37.45      0.68     30.18      1.41      3.17
09:18:45 AM   dev8-34      2.99    742.51      1.60    248.53      0.01      3.40      3.40      1.02
09:18:45 AM   dev8-35      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM   dev8-48     18.16      0.00    645.11     35.52      0.41     22.65      1.09      1.98
09:18:45 AM   dev8-49     18.16      0.00    645.11     35.52      0.41     22.65      1.09      1.98
09:18:45 AM   dev8-50      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:18:45 AM   dev8-51      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

saída sar -w:

12:00:01 AM   cswch/s
05:10:01 AM  51556.52
05:20:02 AM  54484.96
05:30:01 AM  57533.58
05:40:02 AM  57956.67
05:50:01 AM  57885.02
06:00:02 AM  46017.63
06:10:01 AM  21778.29
06:20:01 AM  43464.68
06:30:02 AM  51766.88
06:40:01 AM  53879.12
06:50:02 AM  56531.08
07:00:02 AM  57732.89
07:10:01 AM  57658.24
07:20:02 AM  58209.16
07:30:01 AM  58446.37
07:40:02 AM  58537.84
07:50:01 AM  58243.39
08:00:02 AM  58362.02
08:10:01 AM  58291.73
08:20:02 AM  58370.15
08:30:01 AM  58545.73
08:40:02 AM  58448.40
08:50:01 AM  58198.55
09:00:02 AM  58313.22
09:10:01 AM  58122.85
09:20:02 AM  58517.96
09:30:01 AM  58338.94
09:40:02 AM  58317.30
09:50:01 AM  58312.40
10:00:01 AM  58337.53
10:10:02 AM  58167.55
10:20:01 AM  58408.61
10:30:01 AM  58133.29
10:40:01 AM  58165.08
10:50:02 AM  58240.77
11:00:01 AM  58236.16

Average:      cswch/s
Average:     55991.47

saída sar -I SUM:

12:00:01 AM      INTR    intr/s
05:10:01 AM       sum   3825.03
05:20:02 AM       sum   3999.32
05:30:01 AM       sum   4038.10
05:40:02 AM       sum   4041.99
05:50:01 AM       sum   4015.39
06:00:02 AM       sum   3450.48
06:10:01 AM       sum   2385.73
06:20:01 AM       sum   3355.96
06:30:02 AM       sum   3641.50
06:40:01 AM       sum   3807.91
06:50:02 AM       sum   3853.99
07:00:02 AM       sum   3951.29
07:10:01 AM       sum   3996.06
07:20:02 AM       sum   4005.63
07:30:01 AM       sum   3939.43
07:40:02 AM       sum   3901.39
07:50:01 AM       sum   3920.22
08:00:02 AM       sum   3950.27
08:10:01 AM       sum   3926.09
08:20:02 AM       sum   4072.29
08:30:01 AM       sum   4058.93
08:40:02 AM       sum   3994.94
08:50:01 AM       sum   3969.04
09:00:02 AM       sum   3976.33
09:10:01 AM       sum   3904.43
09:20:02 AM       sum   4054.35
09:30:01 AM       sum   4006.06
09:40:02 AM       sum   3962.65
09:50:01 AM       sum   4016.83
10:00:01 AM       sum   4064.52
10:10:02 AM       sum   3934.29
10:20:01 AM       sum   4029.60
10:30:01 AM       sum   3939.23
10:40:01 AM       sum   3937.37
10:50:02 AM       sum   3961.87
11:00:01 AM       sum   4014.83

Average:         INTR    intr/s
Average:          sum   3794.24

Alguma idéia do que poderia ser, ou onde eu deveria procurar mais informações de diagnóstico?

Obrigado.

linux  high-load 
— SyRenity
fonte
adicionar saída "ps aux"
— alvosu
Saída foi adicionada.
— SyRenity

Respostas:

7

Intuitivamente, eu suspeito que um problema de disco seja a causa mais direta, mas isso não significa que seus discos estejam muito lentos. Sua% iowait do iostat não indica que nenhum processo do usuário esteja gastando muito tempo aguardando a E / S do disco. No entanto, o tempo de CPU no kswapd me preocupa:

root       493  0.1  0.0      0     0 ?        S<    2010  94:48 [kswapd1]

Os 242MB de swap que você está usando podem não parecer muito, mas para atingir esse tipo de tempo de CPU em um sistema que está esgotado há 42 dias, você tem muita atividade de swap acontecendo ou está demorando uma eternidade para terminar uma vez iniciado por causa de outra contenção de disco. Se essa é a fonte do seu problema ou não, é algo que eu definitivamente analisaria.

Você pode executar sar -We publicar as estatísticas de troca para o seu sistema?

— jgoldschrafe
fonte
resultados sar -W adicionados à pergunta acima.
— SyRenity
foi porque você estava usando swap? Veja também serverfault.com/a/524818/27813
— rogerdpack
2

A causa mais comum de carga alta são as unidades lentas. Tente executar o seguinte

sar -d 5 0

e olhando para o campo% util. Se esse número for superior a 70% para qualquer uma das unidades, essa unidade será lenta no processamento de solicitações de E / S que causam alta carga.

Editar: pode funcionar bem em 70%, mas é nesse ponto que você provavelmente começará a ver a degradação do desempenho. Quanto mais alto você for, pior será.

— Patrick
fonte
1
Veja o tamanho da fila e os tempos de serviço. Se estes aumentarem, você tem um problema.
— BillThor
Sim, eles podem ajudar a indicar problemas, mas eu não os mencionei porque os vi disparar sem nenhum efeito adverso.
— 711 Patrick
2
Usar o desempenho do disco como primeira parada em uma análise de causa raiz pode ser perigoso, no entanto. Um dos motivos pelos quais a E / S do disco é um dos últimos lugares em que procurarei um problema, mesmo que sintomático seja a causa de muitos problemas, é que existem muitas coisas engraçadas que podem surgir errado em um sistema que fará com que ele comece a martelar o disco com outras coisas que não a carga de trabalho pretendida, e tentar corrigir o problema adicionando eixos mais ou mais rápidos nem sempre pode ser a melhor abordagem.
— jgoldschrafe
@jgoldschrafe bem, é claro, se o disco for o problema, você terá que descobrir o porquê. Você não pode começar a analisar por que os processos estão mastigando a E / S de disco até confirmar que a alta carga está sendo causada pela E / S de disco.
— 711 Patrick
Publiquei os resultados sar-d, parece muito abaixo do limite de 70%.
— SyRenity
1

Muitos processos do sistema no estado "S <". Na minha máquina, eles são listados apenas como "S". Do homem ps: < high-priority (not nice to other users). Algo parece realmente ferrado. Tente atualizar seu kernel, se possível, e reinicie.

— poige
fonte
"onde eu deveria procurar mais informações de diagnóstico?" - tente pesquisar nos logs, verifique a saída do dmesg.
— poige
Isso é normal para esses processos.
— 711 Patrick
Estou executando a versão 2.6.18-194.el5, algum sentido em atualizar para a versão mais recente disponível?
— SyRenity
Além disso, eu vejo uma morte no OOM que aconteceu na semana passada, não tenho certeza se isso está relacionado.
— SyRenity
Se não me engano, há o kernel-2.6.18-238 em estado selvagem, então certamente tem sentido o tempo todo.
— poige
1

Que tipo de conexão de rede seu servidor possui? Eu já vi cargas subindo rapidamente em situações em que a conexão ao switch deveria ser de 100 Mbit / s full duplex , mas, por algum motivo, foi negociada como 100 Mbit / s half duplex . Depois de forçar o modo 100M-FD com ethtool, as cargas caíram abaixo de 1 e as velocidades de transferência de rede voltaram ao normal.

— Janne Pikkarainen
fonte
O Ethtool mostra uma conexão de Velocidade: 1000Mb / s, Duplex: Cheio.
— SyRenity
OK, provavelmente não é um problema de rede, então. A seguir, você deve verificar: número de alternâncias e interrupções de contexto. Faça isso com sar -w (para opções de contexto) e sar -I SUM (para interrupções).
— Janne Pikkarainen
Eu adicionei as duas saídas na descrição acima.
— SyRenity
As alternâncias de contexto estão em um lado bastante alto; Normalmente, vejo algo entre centenas e milhares de switches / s, mesmo em servidores carregados. Talvez esses processos Java estejam fazendo algo estranho?
— Janne Pikkarainen
Os aplicativos Java são altamente multithread, talvez essa seja a causa da alternância de contexto?
— SyRenity
-1

Você escreveu que vê uma morte de OOM em seus logs. Depois que o assassino do OOM é acionado, você deve reiniciar o mais rápido possível.

Então reinicie e seus problemas desaparecerão.

Richard

— rems
fonte
3
A interrupção do OOM removeu o processo ofensivo que ocupa a maior parte da memória; portanto, por que uma reinicialização é necessária?
— SyRenity
Ao utilizar nosso site, você reconhece que leu e compreendeu nossa Política de Cookies e nossa Política de Privacidade.
Licensed under cc by-sa 3.0 with attribution required.