Every ten minutes the performance counters blank out.
Is this caused by the memory scanning?
The machine is also monitored with veeam monitor, but I have not noticed this behavior on any of the other machines I monitor.
My interim step was to stick all the machines in a resource pool and put hte total limit of teh resource limit to 2 GHZ less than the physical cpu limit.
hmmm, after twaking values, it appears the issue was caused by something on the vi client.
Rebooting the machine running the vi client resolved the issue.
I have determined what is causing the problem.
Every ten minutes on esxi 3.5 it runs this script:
/sbin/decodeSel.sh which collects the ipmi data.
This is causing the performance data blanks.
what hardware is this on, for interest? And also which image (standard/hp/dell/ibm)?
blah,
So far my hypothesis has been soundly disproved:
Namedly the ipmi collection is causing the blank.
So i edited /var/spool/crontabs/root and commented out that collector line, then kill crond and restarted it
ps ax|grep crond.; kill pidnumber; crond -p
and the problem did not go away.
Now I think it is caused by the memory scanning.