Vmware ESX 3.5 Host CPU % wait

[RAINBOW007]

Hello Friends,

I have noticed some CPU statistic for when i run esxtop command and see quite a few readings under %WAIT for the CPU statistics. Can someone explain me what it is and does this mean My VM's are running under poor performance? What should i do?

<table>

3:32:08pm up 492 days 21:42, 124 worlds; CPU load average: 0.03, 0.04, 0.03

PCPU(%):   3.31,   5.19,   3.18,   1.69,   4.34,   4.76,   2.84,   3.16 ;   used total:   3.56

CCPU(%):   0 us,   0 sy, 100 id,   0 wa ;       cs/sec:    480

     ID    GID NAME             NWLD   %USED    %RUN    %SYS   %WAIT    %RDY   %IDLE  %OVRLP   %CSTP  %MLMTD

      1      1 idle                8  800.00  800.00    0.00    0.00   27.29    0.00    0.26    0.00    0.00

      2      2 system              6    0.01    0.01    0.00  600.00    0.00    0.00    0.00    0.00    0.00

      6      6 helper             23    0.01    0.01    0.00 2300.00    0.01    0.00    0.00    0.00    0.00

      7      7 drivers            28    0.09    0.33    0.41 2800.00    0.13    0.00    0.03    0.00    0.00

      8      8 vmotion             1    0.00    0.00    0.00  100.00    0.00    0.00    0.00    0.00    0.00

      9      9 console             1    1.35    1.32    0.03  100.00    0.48  117.85    0.07    0.00    0.00

     15     15 vmware-vmkauthd     1    0.00    0.00    0.00  100.00    0.00    0.00    0.00    0.00    0.00

    187    187 SvrWEBSERVER1       7    2.80    2.80    0.02  700.00    0.74  235.81    0.30    0.00    0.00

    219    219 REPORTS             7    6.32    6.33    0.02  700.00    1.36  231.71    0.41    0.00    0.00

   1216   1216 TNSEXCHANGE         7    2.49    2.49    0.01  700.00    0.66  236.13    0.23    0.00    0.00

   1422   1422 VCS1                5    7.82    7.21    0.66  500.00    1.02  112.05    0.25    0.00    0.00

   1445   1445 vtrain3             5    0.86    0.86    0.00  500.00    0.15  118.91    0.06    0.00    0.00

   1455   1455 vtrain1             5    0.74    0.74    0.00  500.00    0.22  119.19    0.14    0.00    0.00

   1468   1468 ALPH4               5    1.14    1.14    0.00  500.00    0.10  118.60    0.08    0.00    0.00

   1478   1478 ALPH5               5    1.24    1.24    0.00  500.00    0.31  118.77    0.07    0.00    0.00

   1482   1482 vnewrec2            5    0.86    0.86    0.00  500.00    0.20  118.83    0.11    0.00    0.00

   1483   1483 ALPH2               5    0.97    0.97    0.00  500.00    0.24  119.09    0.33    0.00    0.00

</table>

[djciaro]

The definition: Percentage of time the resource pool, virtual machine, or world spent in the blocked or busy wait state. This percentage includes the percentage of time the resource pool, virtual machine, or world was idle.

What’s does this mean:

CPU Wait happens when the CPU cannot perform computations as it is waiting on I/O operations.  This usually happens because of disk or network I/O operations (normally the slowest parts of a system)  for example The logon/logoff process of terminal services could generate a good lot of disk I/O (and sometimes network I/O), That means its normal that your CPU spends an increased amount of time waiting on these operations.

Another time that can cause %WAIT is when your HOST is swapping due storage of memory, have a look and see if this is the case.

wait time is normally just not when its excessive

use guest level monitoring to see if there is anytime consuming too  uch I/O

e.g. perfmon (win) and top (Linux)