I was hoping to get a little insight as to why my %wait times seems exteremely high. Below is a ESXtop extract of my ESX box currently running. A copy/paste into notepad makes it much more readable. Any comments would be appreciated. thanks
login as: pete
pete@172.27.100.204's password:
\[pete@st-demovm pete]$ esxtop
esxtop: Need to run as user root
\[pete@st-demovm pete]$ su -
Password:
\[root@st-demovm root]# esxtop
5:40:02am up 18:14, 82 worlds; CPU load average: 0.04, 0.04, 0.06
PCPU(%): 8.18, 2.36, 4.92, 2.48, 2.04, 0.64, 0.50, 1.97 ; used total: 2.88
LCPU(%): 7.85, 0.34, 1.47, 0.89, 2.70, 2.22, 1.29, 1.18, 0.35, 1.68, 0.43, 0.21, 0.24, 0.26, 0.80, 1.17
ID GID NAME NMEM %USED %SYS %OVRLP %RUN %WAIT %BWAIT %TWAIT %CRUN %CSTP %IDLE %RDY
1 1 idle 16 777.62 0.00 0.02 22.96 0.00 0.00 0.00 0.01 0.00 0.00 800.00
2 2 system 5 0.05 0.00 0.00 0.06 500.00 0.00 500.00 0.00 0.00 0.00 0.00
6 6 console 1 5.30 0.01 0.04 5.57 47.70 46.42 94.12 0.00 0.00 94.09 0.46
7 7 helper 13 0.02 0.00 0.00 0.01 1300.00 0.00 1300.00 0.00 0.00 0.00 0.00
8 8 drivers 7 0.00 0.00 0.00 0.01 700.00 0.00 700.00 0.00 0.00 0.00 0.00
12 12 vmware-vmkauthd 1 0.00 0.00 0.00 0.00 100.00 0.00 100.00 0.00 0.00 0.00 0.00
13 13 TDS-demodrive-T 7 1.13 0.00 0.20 1.52 428.88 270.42 699.30 0.00 0.00 99.02 0.05
14 14 9.3.0.1-Merge-d 5 1.68 0.00 0.16 2.11 288.89 208.79 497.68 0.00 0.00 98.04 0.30
20 20 9.3.0.1-Merge-d 5 1.54 0.00 0.18 1.36 264.05 234.59 498.63 0.00 0.00 98.81 0.07
25 25 Oracle-demodriv 5 1.95 0.00 0.11 2.20 365.23 132.45 497.68 0.00 0.00 97.95 0.19
26 26 OBIE 7 6.22 0.02 0.39 9.79 390.07 298.71 688.79 0.00 0.00 88.83 1.49
28 28 TDS-demodrive-m 5 2.53 0.05 0.11 2.71 354.28 142.98 497.26 0.00 0.00 97.50 0.09
32 32 AB2-demodrive 5 1.65 0.00 0.16 1.99 281.72 216.32 498.04 0.00 0.00 98.19 0.03
You may want to expand single VMs into their sub processes (Press 'e' in esxtop), so you get an idea of what sub processes are waiting all the time.
\- Andreas
here is a more detailed look at a single VM, are these levels normal, I'm a little confused as to why vcpu0 is being reported on this VM and shows such a high level
ID GID NAME NMEM %USED %SYS %OVRLP %RUN %WAIT %BWAIT %TWAIT %CRUN %CSTP %IDLE %RDY
5 2.35 0.00 0.13 2.28 337.64 160.37 498.00 0.00 0.00 96.75 0.10
1158 34 vmware-vmx 1 0.05 0.00 0.00 0.08 97.36 2.62 99.99 0.00 0.00 0.00 0.01
1159 34 vmm0:AB2-demodr 1 2.01 0.02 0.06 1.83 34.60 63.60 98.20 0.00 0.00 97.97 0.04
1160 34 vmware-vmx 1 0.00 0.00 0.00 0.00 100.00 0.06 100.00 0.00 0.00 0.00 0.00
1161 34 mks:AB2-demodri 1 0.23 0.00 0.06 0.40 37.14 62.47 99.61 0.00 0.00 0.00 0.06
1162 34 vcpu-0:AB2-demo 1 0.02 0.00 0.00 0.01 100.00 0.01 100.00 0.00 0.00 0.00 0.00
Hi,
the high wait times you are seeing are perfectly normal because the %TWAIT value includes the %IDLE value. If you substract the idle time from twait time you get the time that the process/VM was really waiting for an event (s. "man esxtop").
Also %TWAIT looks like the sum of %WAIT and %BWAIT. However, I do not really know what the difference is between the three. Anyone?
The value to care most about is %RDY (the ready time). It shows how often a VM was ready to run, but could not get any resources to actually run. So, ready times should be low.
There is an interesting white paper about this: Ready Time Observations[/url].
Hint: please put the esxtop-output into \[ code ] tags (without the blanks) in your posts. This will make it much better readable.
\- Andreas
Message was edited by:
peetz
to add the code-tag hint